SEARCH

SEARCH BY CITATION

Keywords:

  • Age-related memory impairment;
  • aging;
  • anti-ROS system;
  • Drosophila;
  • reactive oxygen species

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Damage from reactive oxygen species (ROS) is thought to be a cause of organismal aging. Reactive oxygen species have also been proposed to be responsible for several age-associated phenotypes, including age-related memory impairment (AMI). However, it has not previously been tested whether increasing ROS affects AMI onset. Here we examined the effects of feeding hydrogen peroxide, and the ROS-generating agent, paraquat, on olfactory aversive memory in Drosophila at young ages and during AMI onset. Reactive oxygen species feeding greatly reduced fly survival, and increased oxidized proteins and transcripts of an antioxidant enzyme, catalase (Cat) and a stress-responsive chaperone, heat-shock protein 22 (Hsp22) in fly heads. However, feeding did not impair memory in young wild-type flies, nor did it exacerbate the memory deficits in flies at the onset of AMI. Strikingly ROS feeding did disrupt memory at young ages and accelerated AMI onset was observed when expression of genes involved in the defense system to ROS, including antioxidant enzymes and Hsp22, was reduced in the mushroom bodies, neural centers required for olfactory memory. These results implicate that although ROS production increases upon aging, neuronal functions required for memory processes are sufficiently protected by the defense system to ROS even at the age of AMI onset. Thus we propose that ROS production does not affect AMI onset in Drosophila.

Reactive oxygen species (ROS) including superoxide, hydrogen peroxide (H2O2) and hydroxyl radicals are produced during cellular energy production. Reactive oxygen species damage cellular macromolecules such as nucleic acids, lipids and proteins, thereby disrupting cellular functions. Since generation of ROS is thought to increase upon aging, a free radical hypothesis has been proposed that cumulative oxidative damage produced by ROS causes aging processes and limits life span (Balaban et al. 2005; Beckman & Ames 1998; Harman 1956). However, due to a defense system against ROS (anti-ROS system), ROS are normally scavenged by antioxidant enzymes such as superoxide dismutase (Sod), catalase (Cat) and glutathione peroxidase (Balaban et al. 2005), and damaged proteins are thought to be repaired by stress-responsive molecular chaperones such as heat-shock proteins (Hsps) (Parsell & Lindquist 1993). Thus, the free radical hypothesis predicts that the normal anti-ROS system is not fully sufficient to counteract the deleterious effects caused by ROS in aged organisms. In support of this idea, overexpressing sod1 and sod2 (Parkes et al. 1998; Sun et al. 2002) and feeding Sod/Cat mimetics (Melov et al. 2000) extend the life span of Drosophila and Caenorhabditis elegans. In Drosophila, lack of hsp22 shortens life span (Morrow et al. 2004a), while overexpression of hsp22 extends life span and increases resistance to oxidative damage (Morrow et al. 2004b).

Although it has not been proven that factors affecting life span also affect age-related memory impairment (AMI) (Burger et al. 2010; Horiuchi & Saitoe 2005), prevailing models propose that AMI is also caused by ROS (Bishop et al. 2010). In mammals, age-related accumulation of oxidative damage is observed in the brain (Lu et al. 2004; Murali & Panneerselvam 2007b; Smith et al. 1991), and genetic and pharmacological interventions that enhance activity of the anti-ROS system improve memory of old animals (Hu et al. 2006; Levin et al. 2005; Liu et al. 2003). AMI is also observed in Drosophila, which has a life span of 30–40 days. Using a Pavlovian olfactory aversive association task (Tully & Quinn 1985), we previously showed that AMI in Drosophila appears at 15 days after eclosion, and consists of a significant and specific reduction in middle-term memory (MTM) which can be measured as memory 1 h after single-cycle training (Tamura et al. 2003). However, it is not clear whether ROS are involved in induction of AMI in Drosophila and to what extent the anti-ROS system counteracts the deleterious effects of ROS on memory.

To address these issues, we fed flies with ROS at young ages and at the age of AMI onset. Our results show that neuronal functions required for memory formation are highly protected from ROS by the anti-ROS system. We provide data implicating that although amounts of ROS increase upon aging, ROS seem not contribute to AMI onset due to a robust anti-ROS system.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Fly stocks

UAS-hsp22-IR, UAS-cat-IR and UAS-sod1-IR lines were obtained from the Vienna Drosophila RNAi Center (Vienna, Austria). Other lines were previously described (Yamazaki et al. 2007). All fly lines used in this study except UAS-cat-IR and UAS-sod1-IR lines were outcrossed to our wild-type control line, w(CS10) (Tamura et al. 2003), for at least six generation before use. Flies were raised under a 12/12 h, light/dark cycle at 25°C and 60% humidity.

Feeding of PQ or H2O2

For acute feeding of paraquat (PQ) or H2O2, approximately 100 flies were transferred into vials containing Whatman 3M filter strips soaked with 500 µl of 5% sucrose and indicated concentrations of PQ (methyl viologen; Sigma, St. Louis, MO, USA) or H2O2 (Wako, Osaka, Japan). Mock-treated flies were fed 5% sucrose alone. Flies were then transferred back to standard medium after acute feeding. For chronic feeding of PQ or H2O2, approximately 100 flies were fed instant medium [Formula 4-24 (R); Carolina biological supply company, Burlington, NC, USA] containing indicated concentrations of PQ or H2O2 and transferred to new medium every other day. Controls were treated identically except PQ and H2O2 was not added.

Determination of survival rate after exposure to ROS

Survival rate was calculated by dividing the number of surviving flies at a particular time point after ROS feeding by the total number of flies used in the experiment. Survival of male and female flies was calculated separately, as females were slightly more sensitive to ROS than males. Data are shown as average survival rates.

Quantification of transcripts

Total RNA from Drosophila heads or whole bodies was extracted with TRIzol reagent (Invitrogen, Carlsbad, CA, USA), and cDNA was synthesized using RevTra Ace (TOYOBO, Osaka, Japan). cDNAs were then analyzed by quantitative real-time polymerase chain reaction (PCR) using an Applied Biosystems model 7500 machine (Applied Biosystems Inc, Foster City, CA, USA) or conventional PCR. rp49 transcripts were used for normalization. Primer sequences used were as follows: 5′-GTCCTGACCATCAGTGTGC-3′ and 5′-CCAGTCTGCTC GATGGTCAC-3′ for hsp22, 5′-CTGATTCCTGTGGGCAAAAT-3′ and 5′-AGGAGAACAGACGACCATGC-3′ for cat, 5′-CACGGTTTTCTTCGA ACAGG-3′ and 5′-CATTGGTGTTGTCACCGAAC-3′ for sod1, 5′-GCCC GTAAAATTTCGCAAAC-3′ and 5′-TCTCCCGGCAGATGATAG-3′ for sod2 and 5′-CACCGGATTCAAGAAGTTCC-3′ and 5′-GACAATCTCCTT GCGCTTCT-3′ for rp49.

Immunohistochemistry

To visualize protein carbonyls, fly heads without proboscides were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 30 min at room temperature. Brains were dissected, and then subjected to whole-mount immunohistochemistry using a Protein Carbonyls Immunohistochemical Staining Kit (COSMO BIO, Tokyo, Japan) according to the manufacturer's instructions. Briefly, fixed brains were treated with PBS containing 0.3% Triton X-100 (PBST) for 1 h, and incubated with a 2,4-Dinitrophenylhydrazine (DNPH) solution for 30 min. DNPH was removed by washing (manufacturer's instructions) and brains were next blocked for 1 h with PBST containing 4% Block Ace (Yukijirushi, Tokyo, Japan), and incubated with a rabbit anti-DNPH antibody diluted at 1:500 in the blocking solution for 1 day at 4°C. A donkey anti-rabbit immunoglobulin G (IgG) Alexa Fluor 488 (Invitrogen) antibody diluted at 1:500 in blocking solution was used for the secondary antibody. Nuclei were stained by incubating brains in TO-PRO™-3 iodide (Invitrogen) diluted at 1:500 in PBST for 45 min. Brains were mounted on Perma Flour (Thermo Fisher Scientific, Waltham, MA, USA), and images were captured using a FluoView 500 confocal microscope (Olympus, Tokyo, Japan). Using similarly prepared fixed brains, TdT-mediated dUTP nick end labeling (TUNEL) assay was performed using the DeadEnd™ Fluorometric TUNEL System (Promega, Madison, WI, USA), according to the manufacture's instructions. As a positive control, a dissected brain was treated with 5 U of DNase (RQ1 RNase-Free DNase, Promega) for 15 min at room temperature, and subjected to TUNEL assay. Nuclei were also stained, and images were captured as described above.

Single-cycle training and memory assays

Single-cycle training and testing of a Pavlovian olfactory association were performed as previously described (Tamura et al. 2003; Tully & Quinn 1985). Briefly, approximately 100 flies were placed in a training chamber where they were exposed to odors and electrical shocks. In single-cycle training, one of the aversive odors, OCT (3-octanol, Sigma) or MCH (4-methylcyclohexanol, Sigma), was paired with electrical shocks (60V 1.5-second pulses every 5 seconds), whereas the other was not. Testing was performed for 1 h after the training by placing flies at a choice point between the two odors for 1.5 min. A performance index (PI) was calculated so that a 50:50 distribution (no memory) yielded a PI of zero and a 0:100 distribution away from the shock-paired odor yielded a PI of 100. Individual PIs were always the average of two experiments where the shock-paired odor was alternated.

Sensorimotor responses to odors and electrical shocks

Peripheral control experiments including avoidance to odors and electrical shocks were performed as previously described (Tamura et al. 2003; Tully & Quinn 1985). Approximately 100 naÏve flies were placed at a choice point of a T-maze where they had to choose between an odor (OCT or MCH) and mineral oil (avoidance to odors) or between electrically shocked and non-shocked conditions (avoidance to electrical shocks). A PI was calculated as described above.

Statistics

All data are expressed as the mean ± SEM. Statistical analyses were performed by using Prism version 5.01 (Graphpad Software, La Jolla, CA, USA). P values <0.05 were judged as statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Aging and PQ feeding increase the expression of hsp22 and catalase in fly heads

Accumulation of oxidized proteins is observed from 14 days after eclosion (Sohal et al. 1993) and, an increase in expression of the oxidative stress response gene, hsp22 (Landis et al. 2004) in fly heads, which is used to monitor ROS levels (King & Tower 1999; Landis et al. 2004; Yang & Tower 2009), is observed after 12 days of age (King & Tower 1999). Given that the onset of AMI is observed at 15 days of age in our wild-type flies (Tamura et al. 2003): these previous findings suggest that amounts of ROS are increased at the age of AMI onset. Therefore, we first examined whether the expression of hsp22 as well as antioxidant enzymes, cat, sod1 and sod2, is increased in fly heads, (1) at 15 days of age, and (2) at younger ages after acute feeding with PQ, a superoxide-generating (redox-cycling) agent (Bus & Gibson 1984). In addition to increases in hsp22 expression, we found significant increases in cat expression in fly heads at 15 days of age (Fig. 1a) and after feeding PQ (Fig. 1b). These results suggest that ROS amounts in fly heads are increased by the age of AMI onset. They also suggest that PQ feeding can be used to artificially increase ROS amounts in fly heads.

image

Figure 1. Transcription of genes associated with the response to ROS is upregulated at AMI onset. (a) Expression of hsp22, cat, sod1 and sod2 in fly heads at the onset of AMI. Flies were reared on standard medium for 3 days or 15 days after eclosion. Expression of each gene was normalized to rp49 expression. Relative expression to 3-day-old was indicated. Expressions of hsp22 and cat were increased in 15-day-old fly heads (P < 0.03 for hsp22 and P < 0.05 for cat by t-test). N = 3 for all data. (b) Expression of hsp22, cat, sod1 or sod2 after feeding with PQ. Two-day-old flies were fed 5% sucrose with or without 10 mm PQ for 1 day. Expressions of hsp22 and cat were increased by PQ feeding (P < 0.001 for hsp22 and P < 0.01 for cat by t-test). N = 3 for all data.

Download figure to PowerPoint

Acute increases in ROS do not affect the memory in young flies or in flies at the age of AMI onset

If an age-related increase in ROS causes AMI, artificially increasing ROS should induce memory defects similar to AMI at young ages and exacerbate memory defects at the age of AMI onset. To address these possibilities, we examined whether feeding young and old flies with PQ or H2O2 impairs 1-h memory, the memory component disrupted in AMI after aversive olfactory training (Tamura et al. 2003).

In order to determine the optimal concentrations of PQ and H2O2 for feeding, we first monitored the survival rate of flies after feeding them PQ or H2O2 for 1 day. Survival rates were decreased to 60–70% by feeding young (3-day-old) flies with 10 mm PQ or 10% H2O2 (Fig. 2a), and by feeding aged (15-day-old) flies with 5 mm PQ or 5% H2O2 for 1 day (Fig. 2b). However, these treatments did not affect memory of young and aged flies, respectively (Fig. 2c). We also fed 3-day-old flies with PQ at a higher concentration, 30 mm, which reduced the survival rate to less than 50%, but this also had no effect on 1-h memory. Significantly, due to AMI onset, mock-treated aged flies showed a slight but significant decrease in 1-h memory compared to young flies. However, 1-day feeding of 5 mm PQ or 5% H2O2 did not further disrupt 1-h memory in aged flies.

image

Figure 2. Acute exposure to ROS does not affect olfactory memory in young flies and the flies at AMI onset. (a, b) Survival after acute feeding of PQ or H2O2. Two-day-old flies (a) or 15-day-old flies (b) were fed 5% sucrose with the indicated concentrations of PQ or H2O2 for 1 day (indicated by black bars). Data is shown as average survival rates of males and females. (c) Olfactory memory after acute feeding of PQ or H2O2. Three-day-old flies or 15-day-old flies were fed with PQ or H2O2 for 1 day and 1-h memory was measured. Two-way analysis of variance indicates significant effects of age (P < 0.001) but not treatment (P = 0.86) or interaction between age and treatment (P = 0.575). Mock-treated 15-day-old flies showed significant reduction in 1-h memory compared to 3-day-old flies due to AMI (***P < 0.001 by Bonferroni-Dunn post hoc analyses). N = 6–8. (d) Relative expression of hsp22 after acute feeding of PQ or H2O2. Three-day-old flies or 15-day-old flies were treated as in (c). N = 3 for all data. (e) Immunohistochemistry for protein carbonyls. Three-day-old flies were fed with or without 10 mm PQ for 1 day. Protein carbonyls (green) and nuclei (red) were visualized in dissected brains. Approximate location of Kenyon cells (KCs) and calyces of the MBs are shown in the middle and the right panels. Scale bars represent 50 µm.

Download figure to PowerPoint

Increases in ROS amounts in fly heads after feeding either PQ or H2O2 were examined by measuring hsp22 transcripts in fly heads (Fig. 2d). In young flies, hsp22 expression in fly heads was increased after 1-day feeding of 10 mm PQ or 10% H2O2. These increases were higher than or equivalent to the increase seen in untreated aged flies. Significantly, hsp22 expression was further elevated in 15-day-old fly heads after 1-day feeding of 5 mm PQ or 5% H2O2. To verify whether these ROS-dependent changes in hsp22 expression are associated with oxidative damage in neuronal structures important for memory, we monitored oxidized proteins by staining for protein carbonyls (Stadtman 1992) after 1-day feeding of PQ, in the Kenyon cells of the mushroom bodies (MBs), brain structures critical for memory (Heisenberg 2003) and AMI (Yamazaki et al. 2007). Protein carbonyls were highly enriched in various regions including the Kenyon cells after 1-day feeding of 10 mm PQ (Fig. 2e), demonstrating that oxidative damage in the brain is induced by feeding PQ. Taken together, these results show that acute exposure to ROS does not impair memory at young ages or at the age of AMI onset and implicate that AMI onset is not caused by an age-related increase in ROS.

Chronic exposure to ROS does not accelerate AMI

The results obtained above left the possibility that AMI is caused specifically by long-term exposure to ROS rather than acute exposure. If this is the case, chronic exposure to exogenous ROS should exacerbate AMI at 15 days of age. Chronic feeding of 2 mm PQ or 0.3% H2O2 to flies from 3-day-old to 15-day-old resulted in approximately 50% and 30% decreases in viability, respectively (Fig. 3a). However, chronic feeding of PQ or H2O2 did not cause a decrease in 1-h memory at 15 days of age (Fig. 3b). hsp22 transcripts in head extracts increased upon feeding PQ or H2O2 for 12 days (Fig. 3c), indicating that these treatments are sufficient to increase amounts of ROS in fly heads during aging. Thus, these results further suggest that AMI onset is not caused by ROS.

image

Figure 3. Chronic exposure to ROS does not affect olfactory memory. (a) Survival after chronic feeding of PQ or H2O2. Flies were reared in food containing indicated concentrations of PQ or H2O2 from 3 days of age and survival rate was measured until 15 days of age. (b) Olfactory memory after chronic feeding of PQ or H2O2. One-way analysis of variance indicates no effects of treatment on memory (P = 0.59). N = 6–8. (c) Relative expression of hsp22 after chronic feeding of PQ or H2O2. Fifteen-day-old flies reared in PQ or H2O2 were collected and relative expression of hsp22 in heads was determined. One-way analysis of variance indicates significant effects of treatment on hsp22 expression (P < 0.01). N = 3 for all data.

Download figure to PowerPoint

Age-related increases in ROS do not contribute to induction of AMI onset due to the anti-ROS system

Since neither acute nor chronic increases in ROS affect memory at 3 days and 15 days of age, we reasoned that neural functions required for memory may be highly protected from ROS, even at the age of AMI onset, by the anti-ROS system. Alternatively, ROS may not directly affect memory formation. If the former is the case, ROS should disrupt memory when the defense system for ROS is attenuated, while in the later case, ROS should not affect memory even under that condition. The anti-ROS system consists of antioxidant enzymes, including as Cat and Sod1, which directly scavenge ROS (Balaban et al. 2005), and stress response proteins, such as Hsp22, which are thought to function as molecular chaperones that repair damaged proteins. To address the relationship between ROS and the anti-ROS system in memory and AMI, we knocked-down hsp22, cat and sod1 by RNA interference (RNAi) in the MBs.

Inverted repeats (IR) of the hsp22, cat and sod1 transcripts (Dietzl et al. 2007) were expressed by using the GAL4-UAS transgene system (Brand & Perrimon 1993). We confirmed RNAi-dependent reduction of each transcript by using a ubiquitous GAL4 driver, Act5C-GAL4 (Fig. 4). Transcripts of hsp22 were reduced by hsp22-IR expression in Act5C-GAL4>UAS-hsp22IR flies (Fig. 4a). Furthermore, the increase in hsp22 expression after 1-day feeding of PQ was suppressed in Act5C-GAL4>UAS-hsp22IR flies. We also confirmed that cat-IR and sod1-IR driven by Act5C-GAL4 significantly reduced the transcripts of cat and sod1, respectively (Fig. 4b,c). These results indicate that the expression of each IR efficiently reduces the expression of its target gene.

image

Figure 4. RNAi-mediated silencing of hsp22, cat or sod1. (a) Transgenic flies carrying Act5C-GAL4 and UAS-RNAi transgenes for hsp22 (Act5C-GAL4>UAS-hsp22-IR) displayed reduced expression of hsp22 and further reduction of hsp22 expression upon PQ feeding. In the left four lanes, two-fold serial dilutions of the control cDNA prepared from wild-type flies were used, which indicates the linear range of PCR. (b, c) Similar to Act5C-GAL4>UAS-hsp22-IR transgenic flies, Act5C-GAL4>UAS-cat-IR (b) and Act5C-GAL4>UAS-sod1-IR transgenic flies (c) displayed reduced expression of cat and sod1, respectively.

Download figure to PowerPoint

We targeted RNAi expression to the MBs by using the OK107 GAL4 driver (Lee et al. 1999; Yamazaki et al. 2007) and measured 1-h memory at 3 days of age (Fig. 5a). In the absence of PQ feeding, all MB-RNAi transgenic flies had memory similar to control lines. However, after 1-day feeding of PQ, these MB-RNAi flies showed significant reduction in 1-h memory. One-day feeding of PQ did not affect 1-h memory in transgenic flies carrying the GAL4 driver or the UAS transgene alone. Sensorimotor responses to odors and electrical shocks used for training were not significantly affected by 1-day feeding of PQ in MB-RNAi transgenic flies (data not shown), indicating that decreased memory was not due to defects in sensorimotor responses. These results show that although ROS cause memory defects, neuronal functions required for memory formation are highly protected from ROS by the anti-ROS system.

image

Figure 5. Reduced expression of hsp22, cat or sod1 results in memory defects after ROS exposure, and accelerated AMI. (a) Olfactory memory in RNAi transgenic flies for hsp22, cat or sod1 after acute PQ feeding. Feeding 30 mm PQ for 1 day significantly impaired 1-h memory of transgenic flies expressing the RNAi constructs for hsp22, cat or sod1 (*P < 0.05 or **P < 0.01 by t-test, PQ vs. mock for each genotype). N = 6 for all data. (b) Age-related memory decline in RNAi transgenic flies. Two-way analysis of variance indicate significant effects of genotype (P < 0.0001), age (P < 0.0001) and interaction between age and genotype (P < 0.05). Bonferroni-Dunn post hoc comparisons indicate significant differences between wild-type and all RNAi transgenic lines, at 10 days and 15 days of age (**P < 0.01 or ***P < 0.001).

Download figure to PowerPoint

We next examined the effects of lowered anti-ROS system on memory at the age of AMI onset in the absence of PQ. Strikingly, all MB-RNAi transgenic flies displayed larger memory defects at the age of AMI onset (15 days of age) compared to wild-type flies (Fig. 5b). In addition, while age-related memory decline only becomes significant at 15 days of age in wild type, MB-RNAi transgenic flies showed significant age-dependent memory declines at 10 days of age. Neither flies containing the UAS transgenes alone nor the GAL4 driver alone exacerbated the age-dependent decrease in 1-h memory (Fig. 6). Transgenic flies expressing the RNAi constructs did not show age-dependent decline in sensorimotor responses to odors and electrical shocks used for training (data not shown). Taken together, these data further support the idea that production of ROS increases upon aging, however, this production does not contribute to AMI onset in wild-type flies due to robust anti-ROS system.

image

Figure 6. AMI was not enhanced in the genetic control line for MB-RNAi transgenic flies. In contrast to MB-RNAi transgenic flies for anti-ROS system, genetic control flies did not show further memory decay at the age of AMI onset. A two-way analysis of variance of the data indicates significant effect of age (P < 0.001), but not genotype (P = 0.82) and in interaction between age and genotype (P = 0.66). N = 6 for all data.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

In Drosophila, AMI is characterized by a specific reduction in MTM, a memory phase that can be measured 1 h after single-cycle training. Age-related memory impairment onset occurs at 15 days of age and shows an increase in severity upon further aging. Although cumulative damage caused by ROS has been implicated in AMI (Levin et al. 2005; Liu et al. 2003; Murali & Panneerselvam 2007a; Nicolle et al. 2001), we provide data suggesting that AMI onset is not induced by ROS in Drosophila. If ROS cause AMI, an increase in ROS should further enhance AMI. However, neither acute nor chronic feeding with PQ and H2O2 enhanced AMI at 15 days of age, although they caused oxidative damage, decreased life span and induced anti-ROS responses.

Our present data implicate that neuronal functions required for MTM formation are highly protected from ROS due to a robust anti-ROS system which includes Cat, Sod1, Sod2 and Hsp22 by the age of AMI onset. While PQ or H2O2 feeding did not affect memory in wild-type flies, RNAi lines with reduced hsp22, cat and sod1 expression had memory defects after PQ and H2O2 feeding at 3 days of age. They also showed age-dependent memory defects at 10 days of age in the absence of PQ and H2O2 feeding, an age prior to wild-type AMI onset. Since expression of hsp22 and cat increase after feeding PQ and H2O2, activity of the anti-ROS system is likely to be upregulated upon increasing ROS even at 15 days of age. Notably, in contrast to 3-day-old RNAi transgenic flies, 10-day-old RNAi transgenic flies displayed memory defects in the absence of PQ and H2O2 feeding. Taken together, these results suggest that the amount of ROS increases in fly heads upon aging. However, this increase fails to induce memory defects at the age of AMI onset due to increased activity of the anti-ROS system.

Our data show that ROS cause memory defects when activity of the anti-ROS system is decreased. In line with this, observed AMI in rodents is accompanied by decreases in the anti-ROS system, including decreases in glutathione concentration (Murali & Panneerselvam 2007a; Zhu et al. 2006) and decreases in activity of Cat (Tian et al. 1998) and Sod (Gupta et al. 1991). In addition, mammalian AMI is ameliorated by enhancing the anti-ROS system by methods including sod overexpression (Hu et al. 2006; Levin et al. 2005) and administration of Sod/Cat mimetics (Liu et al. 2003). In our hands, overexpression of sod1 did not suppress AMI onset in Drosophila (data not shown) indicating that while an age-related increase in ROS is likely to contribute to mammalian AMI, it may not do so to AMI onset in flies. As AMI onset is significantly delayed by reducing Protein kinase A (PKA) activity in the MBs (Yamazaki et al. 2007; Yamazaki et al. 2010), our current results implicate that age-related increase in PKA-dependent signaling plays critical role in the initiation of AMI in a ROS-independent manner.

In the brain of Alzheimer's disease (AD) patients, oxidative damage (Hensley et al. 1995; Pratico 2008; Smith et al. 1991) is increased, while activity of antioxidant enzymes such as glutathione peroxidase, glutathione S-transferase, Cat and Sod (Marcus et al. 1998; Padurariu et al. 2009) are reduced. Similar phenomena have been reported in transgenic mice mimicking the AD pathology (Schuessel et al. 2005; Smith et al. 1998; Sung et al. 2004). Experimental as well as epidemiologic studies show that dietary intake of antioxidants, including (but not limited to) vitamin E, reduces the risk of AD (Pratico 2008; Sung et al. 2004). Therefore ROS may also be involved in memory defects in AD patients.

It is still not clear how ROS disrupt memory formation. Oxidative damage is often linked to neurodegeneration (Floyd 1999). However, feeding PQ did not induce neuronal apoptosis in the brain of RNAi transgenic flies (Hirano unpublished observation). Therefore, neurodegeneration cannot account for memory defects observed in RNAi transgenic flies. Intriguingly Hsp22 is localized in mitochondria (Morrow et al. 2000). As mitochondria play important roles in memory formation (Ben-Shachar & Laifenfeld 2004), Hsp22 may repair damaged mitochondrial proteins required for memory formation.

While our current study suggests that ROS are not involved in the initial onset of AMI, we do not exclude the possibility that ROS are involved in later aspects of AMI including regulation of AMI severity. Notably, although expression of hsp22 increases at the age of AMI onset, expression decreases in fly heads upon further aging (Landis et al. 2004). This suggests that the anti-ROS system might eventually be overwhelmed by ROS leading to increased severity of AMI.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

We thank J. Tower and R. Tanguay for materials. We also thank members in our laboratory for helpful discussion. This work was supported by funding to M.S. from the Ministry of Education, Culture, Sports, Science and Technology (Grand-in-Aid on innovative Areas ‘Systems Molecular Ethology’ and Scientific Research (B) (21300144)), and to Y.H. from JSPS research fellowship for young scientists.