|Home | About | Journals | Submit | Contact Us | Français|
Age-related locomotor impairment (ARLI) is one of the most detrimental changes that occurs during aging. Elderly individuals with ARLI are at increased risks for falls, depression and a number of other co-morbidities. Despite its clinical significance, little is known about the genes that influence ARLI. We consequently performed a forward genetic screen to identify Drosophila strains with delayed ARLI using negative geotaxis as an index of locomotor function. One of the delayed ARLI strains recovered from the screen had a P-element insertion that decreased expression of the insulin signaling gene phosphoinositide-dependent kinase 1 (PDK1) Precise excision of the P-element insertion reverted PDK1 expression and ARLI to the same as control flies, indicating that disruption of PDK1 leads to delayed ARLI. Follow-up studies showed that additional loss of function mutations in PDK1 as well as loss of function alleles of two other insulin signaling genes, Dp110 and Akt (the genes for the catalytic subunit of phosphoinositide 3-kinase and AKT), also forestalled ARLI. Interestingly, only some of the strains with delayed ARLI had elevated resistance to paraquat, indicating that enhanced resistance to this oxidative stressor is not required for preservation of locomotor function across age. Our studies implicate insulin signaling as a key regulator of ARLI in Drosophila.
Aging in humans and other species is characterized by a number of progressive functional declines (Arking, 1998). Age-related locomotor impairments (ARLI) are particularly problematic in this regard. Patients with impaired locomotor abilities have reduced physical activity levels (Pahor et al., 2006), are less independent and require more clinical care (Boyd et al., 2005), are at much greater risk for being institutionalized (von Bonsdorff et al., 2006), and are at greater risk for depression (Braam et al., 2005). Additionally, ARLI is associated with increased risk for falling and skeletal fractures as well as a variety of co-morbidities including coronary artery and cerebrovascular diseases, obesity, diabetes and hypertension (Boyd et al., 2005; Hillsdon et al., 2005; Onder et al., 2005; Volpato et al., 2005; Cesari et al., 2006a; Cesari et al., 2006b). ARLI is therefore a significant clinical challenge for elderly patients and their caregivers.
ARLI occurs in rodents (Forster et al., 1996) and flies (Grotewiel et al., 2005) in addition to humans (Walston et al., 2006). ARLI is associated with age-dependent decreases in muscle mass and walking speed in vertebrates (Arking, 1998; Woo et al., 1999) as well as in Drosophila (Baker, 1976; Sohal, 1985; Rhodenizer et al., 2008). Additionally, oxidative damage might be an underlying cause of ARLI in humans (Nikolic et al., 2005) and flies (Bhandari et al., 2007). These parallels raise the possibility that ARLI might be driven by common mechanisms in invertebrates and mammals.
Our laboratory has developed an assay called Rapid Iterative Negative Geotaxis (RING) for assessing locomotor function in flies (Gargano et al., 2005). Negative geotaxis (i.e. startle-induced climbing) is a legged locomotor behavior that becomes impaired with age due to an age-dependent decrease in climbing speed (Rhodenizer et al., 2008). Studies from our laboratory using RING assays show that flies with mutations in the insulin signaling gene chico, the metabolite transporter gene Indy, and the olfactory receptor gene Or83b have slowed ARLI (Gargano et al., 2005; Martin and Grotewiel, 2006; Rhodenizer et al., 2008). Additional studies from our laboratory show that RNAi-mediated knock-down of SOD2, a major mitochondrial antioxidant, accelerates ARLI (Bhandari et al., 2007). These studies support a role for insulin signaling, metabolism, environmental sensing and oxidative damage in ARLI.
A more complete understanding of the genes and genetic pathways that influence ARLI would greatly facilitate the development of interventions that mitigate the effects of age on locomotor ability. To this end we performed a forward genetic screen for P-element insertions that forestall ARLI in Drosophila. We identified a number of transposon insertion strains with delayed ARLI including a strain with a P-element insertion in the gene phosphoinositide-dependent kinase 1 (PDK1) that reduced expression of the gene’s mRNA. Additional studies with this and other loss of function alleles in PDK1 in conjunction with loss of function alleles in two other insulin signaling genes, Dp110 and Akt, confirmed a role for insulin signaling in ARLI.
All flies for all studies were reared at 25°C and 60% relative humidity under a 12-hour light/dark cycle on standard Drosophila medium (10% sucrose, 2% yeast, 3.3% cornmeal, 1% agar, 0.2% Tegosept) supplemented with active yeast. For all aging studies, adult flies (1–3 days old) were briefly anesthetized with carbon dioxide (CO2) and then placed into fresh food vials at 25 flies per vial. Adult flies were transferred to fresh food vials twice weekly for all ARLI studies. EP lines were obtained from the Drosophila Stock Center (Bloomington, IN, USA). pGawB lines were kindly provided by Laurent Seroude (Queens University, Kingston, ON, Canada). Our standard laboratory stock, w[cs], contains the w1118 allele in the Canton-S genetic background. All EP and pGawB element chromosomes identified in the primary and secondary screens were backcrossed for seven generations to w[cs] prior to testing in the confirmation phase of the screen (see Results). Revertant chromosomes generated by precise excision of EP837 from the PDK1 locus were generated using the Δ2–3(99B) transposase source as described (Stoltzfus et al., 2003).
Negative geotaxis (the bang-induced distance climbed at 4s) was assessed in Rapid Iterative Negative Geotaxis (RING) assays as originally described (Gargano et al., 2005) except that flies that did not perform the behavior were scored as zero. Multiple genotypes of age-matched groups were tested simultaneously in all studies. The primary screen was performed by assessing negative geotaxis in 2 vials of flies per strain tested at 4 weeks of age. The secondary screen was performed by assessing 4 vials of flies per strain at 1 and 4 weeks of age. All other tests were performed on 5 vials of flies per strain per experiment at weekly intervals. Each vial of 25 flies used for RING tests was from an independent food bottle. Analysis of variances and post hoc tests were performed with JMP 5.01 or Prism 4.03.
Confirmation of previously mapped P-element insertions was performed using primers that anneal to genomic DNA flanking the insertion sites and to the P-elements (Table 1) using standard protocols. The locations of previously unmapped PGawB elements were identified through inverse PCR (Dalby et al., 1995). Genomic DNA was isolated from a sample of 50 flies from each pGawB line using a DNeasy Blood and Tissue Kit (Qiagen). Genomic DNA was digested with DpnII or MspI and then self-ligated with T4 DNA ligase (Invitrogen). Ligation products were used as templates for inverse PCR using primers (5’-CCCCACGGACATGCTAAGG-3’, 5’-CTCATAATATTAATTAGACGAAATTATTTTTAAAG-3’) that anneal to the 3’ end of PGawB. Amplification products from inverse PCR reactions (i.e. DNA flanking the pGawB insertions) were sequenced using a primer (5’-CGACACTCAGAATACTATTCCTTTCAC-3’) that anneals near the 3’ end of pGawB.
Gene expression was assessed via quantitative real-time PCR (qRT-PCR) using standard methods. Whole-body RNA was isolated using TRIZOL (Invitrogen), treated with DNAse (Ambion), and reverse transcribed using an oligo(dT) primer and Superscript II reverse transcriptase (Invitrogen). qRT-PCR was performed using an Applied Biosystems Fast 7500 unit. Expression of PDK1, Pfrx, Doc3, m6, HLHm7 and CG14045 was detected using SYBR Green (Quanta Biosciences) with custom primers (Table 2) and normalized to Actin5c expression. For detection of Dp110, a Taqman assay was used (#Dm02142683_g1) with RPII140 (#Dm02134593_g1) as endogenous control. Each quantitative RT-PCR experiment was performed in triplicate and repeated three or more times with independent RNA and cDNA samples.
Males and females were collected for each genotype under brief CO2 anesthesia and given a period of 24 hours to recover. Flies were starved for 6 hours and then placed in vials with Whatman filter disks saturated with 300 µl of 40mM paraquat in 5% sucrose. The vials were placed in a humidified box at room temperature and the number of surviving flies in each vial was recorded after 18 hours. Each paraquat experiment was repeated 3 times with 6 vials of 25 flies per genotype and sex.
We performed a forward genetic screen to identify P-element insertions in Drosophila that delay age-related locomotor impairment (ARLI). We used a Rapid Iterative Negative Geotaxis (RING) assay previously developed by our laboratory (Gargano et al., 2005) to assess ARLI throughout the screen and for all follow-up studies. We tested only males in our screen toward limiting the potential confounding effects of reproduction on aging in female flies (Chapman et al., 1995; Sgro and Partridge, 1999; Partridge et al., 2005). We screened a total of 729 EP (Rorth, 1996) and 364 pGawB (Seroude et al., 2002) transposon insertions. Although EP and pGawB elements have various uses in Drosophila, we focused on their ability as transposons to disrupt the function of a gene when inserted nearby (Stoltzfus et al., 2003).
As a prelude to the screen, we assessed negative geotaxis at 7 and 28 days of age in 20 randomly chosen EP and pGawB lines (Figure S1A and S1B, see Supplemental Material). Negative geotaxis was substantially reduced by 28 days of age in these randomly selected flies, consistent with our previous studies with several other fly strains (Cook-Wiens and Grotewiel, 2002; Goddeeris et al., 2003; Gargano et al., 2005; Martin and Grotewiel, 2006; Bhandari et al., 2007). We therefore screened for strains within the EP and pGawB collections that had preserved negative geotaxis at 28 days of age.
We anticipated that our screen would identify 2 classes of flies. Class 1 flies would have normal negative geotaxis at a young age that would become impaired more slowly than control flies (Figure S1C). Class 2 flies would have elevated negative geotaxis at all ages and would therefore be hyperfunctional regardless of age (Figure S1C). We chose to pursue Class 1 flies in our studies because they are more consistent with delayed ARLI than are Class 2 flies (Martin et al., 2005).
In the primary screen, we assessed negative geotaxis in all EP and pGawB transposon lines at 28 days of age. The data from this phase of the screen generally had Gaussian distributions (Figure S2A and S2B, see Supplemental Material). We selected 100 EP and 30 pGawB transposon insertion lines with the best negative geotaxis performance at 28 days of age (Figure S2C and S2D) for testing in the secondary screen.
In the secondary screen, we assessed negative geotaxis at 7 and 28 days of age in the 130 transposon lines selected from the primary screen. Negative geotaxis in some lines was reduced considerably by 28 days of age, whereas the behavior was well maintained in other lines during this portion of adulthood (representative EP and pGawB lines, Figure S3A and S3B, see Supplemental Material).
We selected the best 24 transposon lines from the secondary screen (i.e. lines with strong negative geotaxis and good survival at 28 days of age) for further testing in the confirmation phase of the screen. To control for genetic background effects on behavior and aging, we backcrossed the 24 lines for 7 generations to w[cs], our standard laboratory stock (Cook-Wiens and Grotewiel, 2002; Goddeeris et al., 2003; Gargano et al., 2005; Bhandari et al., 2006; Bhandari et al., 2007). We then assessed negative geotaxis at weekly intervals in the backcrossed transposon lines and in w[cs] control flies. ARLI (age-related impairment of negative geotaxis) was significantly delayed in 7 of the 24 backcrossed transposon lines in multiple experiments (Figure 1A and 1B). The other backcrossed transposon lines either had normal ARLI (11 lines) or were Class 2 flies (i.e. had enhanced negative geotaxis at all ages, 4 lines), excluding them from the project.
We mapped or confirmed the location of the transposon insertions in all of the strains with delayed ARLI via inverse PCR or standard PCR, respectively (primer information provided in Table 1). EP837 is inserted in the transcription unit of phosphoinositide-dependent kinase 1 (PDK1, a.k.a. CG1210, Figure 2A), a gene within the canonical insulin signaling pathway (Taniguchi et al., 2006). EP1150 is inserted within 6-phosphofructo-2-kinase (Pfrx, a.k.a. CG3400, Figure 2B). Pfrx encodes a key, highly conserved enzyme (6-phosphofructo-2-kinase) involved in gluconeogenesis (Rider et al., 2004). DJ708 is inserted upstream of Dorsocross3 (Doc3, a.k.a. CG5093, Figure 2C), a gene that encodes a T-box transcription factor involved in wingless and decapentapalegic signaling (Reim et al., 2003; Hamaguchi et al., 2004; Reim and Frasch, 2005). Three pGawB transposon insertions (DJ913, DJ1026 and DJ996) are inserted between the genes m6 and HLHm7 (Figure 2D). The m6 locus encodes a 70 amino acid protein of unknown function found only in the Drosophila genus (FLYBASE.BIO.INDIANA.EDU). HLHm7 (a.k.a. E(spl)m7 or m7) encodes a helix-loop-helix transcription factor of the enhancer of split family (Klambt et al., 1989; Ligoxygakis et al., 1999). EP1174 (Figure 2E) is inserted near CG14045 which encodes a predicted GTPase activating protein (GAP) for Rho G proteins (FLYBASE.BIO.INDIANA.EDU). Based on BLAST searches (Altschul et al., 1997), the GAP encoded by CG14045 is found in insects but not other animals.
We performed quantitative real-time PCR (qRT-PCR) analyses to determine whether the EP and pGawB insertions that delayed ARLI were associated with altered expression of nearby genes (primer information provided in Table 2). Compared to w[cs] controls, expression of PDK1 and Pfrx were decreased by 25% in EP837 and 92% in EP1150, respectively (Table 3). In contrast, expression of Doc3 was increased by ~8-fold in DJ708 (Table 3). Thus, EP837 and EP1150 are partial loss of function mutations in PDK1 and Pfrx, respectively, whereas DJ708 is a gain of function allele of Doc3. These three alleles are designated PDK1EP837, PfrxEP1150 and Doc3DJ708 hereafter. We analyzed PDK1EP837 and other insulin signaling mutants as described below. The results of ongoing analysis of PfrxEP1150 and Doc3DJ708 will be provided elsewhere.
Despite repeated efforts, we did not consistently detect a change in expression of CG14045 in EP1174 adult flies. Similarly, we did not find a consistent change in expression of either m6 or HLHm7 in DJ913, DJ996 or DJ1026 adult flies. It is possible that expression of these genes might be altered during a limited number of developmental stages or altered in a subset of tissues. Additional studies are ongoing to link the effects of these transposon insertions to specific genes. Nevertheless, we provide preliminary data on these P-element insertion strains in the interest of fully reporting the results of our screen and because these strains facilitated our investigation of the relationship between paraquat resistance and ARLI (see below).
PDK1 is a component of the canonical insulin signaling pathway (Taniguchi et al., 2006). The delayed ARLI in PDK1EP837 flies (Figure 1A) suggested that insulin signaling might influence this aspect of aging. We explored this possibility by further investigating the role of PDK1 and other insulin signaling genes.
To address whether the EP837 insertion was responsible for the delayed ARLI in PDK1EP837 flies, we assessed negative geotaxis across age in revertant flies (PDK1EP837rv4 and PDK1EP837rv9) with wild type chromosomes derived from precise excision of the EP837 transposon insertion. Precise excision of EP837 returned PDK1 expression to normal in both revertant lines (one sample t tests, n = 5–6, n.s.). ARLI was indistinguishable in w[cs] and both revertant lines, whereas it was significantly delayed in PDK1EP837 flies compared to these three controls (Figure 3A). We further examined the role of EP837 in ARLI by evaluating the DT50 (time required for negative geotaxis to decline to 50% of young flies) and total negative geotaxis (total locomotor function across age) (Martin et al., 2005; Martin and Grotewiel, 2006). PDK1EP837 flies had greater T50 and total negative geotaxis values than w[cs] controls (Figure 3B and 3C), indicating an extension of locomotor function across age that gives rise to greater total function (Martin et al., 2005; Martin and Grotewiel, 2006). These two phenotypes were eliminated relative to w[cs] control flies in both of the revertant lines (Figure 3B and 3C). These results demonstrate the EP837 insertion in PDK1 causes delayed ARLI.
PDK1 functions immediately downstream of phosphatidylinositol 3-kinase (PI3K) and immediately upstream of AKT in the canonical insulin signaling pathway (Taniguchi et al., 2006). Based on the reduced expression of PDK1 in EP837 flies (Table 3), we reasoned that additional partial loss of function alleles in PDK1 and other insulin signaling genes might also delay ARLI. We therefore characterized two additional transposon insertions in PDK1 (EP3553 and BG02759, Figure 2A), two transposon insertions in Dp110 (a.k.a. CG4141, the gene that encodes the catalytic subunit of PI3K, Figure 2F), and a transposon insertion in Akt (a.k.a. CG4006, the gene that encodes AKT, Figure 2G). The insertion sites for all transposons in or near these genes were confirmed by standard PCR. All transposons were subsequently backcrossed for seven generations to w[cs] to control for genetic background effects on behavior and aging. qRT-PCR studies indicated that these additional five transposon insertions cause partial loss of function in their respective genes (Table 3). These strains are designated as PDK1EP3553, PDK1BG02759, Dp110c00368, Dp110e03435, and Aktc02098.
To address whether additional partial loss of function alleles in PDK1 and whether partial loss of function in Dp110 and Akt delay ARLI, we assessed negative geotaxis across age in w[cs] controls and PDK1EP3553, PDK1BG02759, Dp110c00368, Dp110e03435/+, and Aktc02098/+ flies. We evaluated males and females in these studies because mutations in insulin signaling genes have sex-specific effects on life span (Burger and Promislow, 2004). Negative geotaxis across age declined in all genotypes as expected (Figure 4). Negative geotaxis was elevated relative to w[cs] controls in PDK1EP3553 and PDK1BG02759 males (Figure 4A) and females (Figure 4B). Negative geotaxis across age was also elevated in Dp110c00368 males and females and in Dp110e03435/+ females (Figure 4C and 4D). Aktc02098/+ males and females also had enhanced negative geotaxis across age (Figure 4E and 4F). All of these effects support a role for insulin signaling in ARLI.
To further investigate the effects of mutations in PDK1, Dp110 and Akt on ARLI, we evaluated DT50 and total negative geotaxis values derived from the data in Figure 4. Relative to w[cs] controls, DT50 (Figure 5A and 5B) and total negative geotaxis (Figure 5C and 5D) was increased in all males and females with transposon insertions in PDK1, Dp110 and Akt except for Dp110c00368 and Dp110e03435/+ males. The increases in DT50 and total negative geotaxis in PDK1, Dp110 and Akt mutants are consistent with a significant preservation of locomotor function across age in these animals. Together, our studies show that partial loss of function mutations in multiple insulin signaling genes delay ARLI in flies.
Increased resistance to exogenously applied oxidative stress is associated with extended life span in Drosophila (Lin et al., 1998; Tatar et al., 2001; Wang et al., 2004; Walker et al., 2006). To investigate whether the delayed ARLI in the transposon insertion strains used in our studies might be associated with altered oxidative stress resistance, we assessed survival in flies fed paraquat, a strong oxidative stressor (Kirby et al., 2002). Paraquat survival was enhanced in all transposon insertion lines directly recovered from our screen (Figure 6A). In contrast, additional studies with other alleles of PDK1 and alleles of Dp110 and Akt revealed that only PDK1BG02759 males had enhanced paraquat survival compared to w[cs] control flies (Figure 6B). In females, none of the additional PDK1, Dp110 or Akt mutants tested had significantly altered paraquat survival (Figure 6C). Although elevated paraquat survival was observed in many transposon lines with delayed ARLI, enhanced resistance to this oxidative stressor is not required for the preservation of locomotor function across age.
Aging is associated with progressive impairments in a number of physiological and cognitive functions (Arking, 1998). Among the functional deficits that occur during aging, ARLI is particularly significant because it is associated with increased rates of falls, fractures and institutionalization in humans (Boyd et al., 2005; Pahor et al., 2006; von Bonsdorff et al., 2006). Identifying genes and genetic pathways that influence age-related changes in locomotor ability and other functions is critical for developing treatment strategies that mitigate the effects of age on health status. Here, we report the results of our studies in Drosophila that investigate the genetic basis of ARLI.
Through a forward genetic strategy, we identified seven transposon insertion lines with age-dependent preservation of negative geotaxis, a startle-induced locomotor behavior (Rhodenizer et al., 2008). One of the lines with delayed ARLI has a transposon insertion in PDK1 that reduces expression of this gene. Removal of the transposon insertion via precise excision restores normal expression of PDK1 and eliminates the delayed ARLI phenotype in these flies, suggesting that PDK1 and, more generally, insulin signaling influence this aspect of aging. Consistent with this possibility, ARLI is delayed in flies with other partial loss of function alleles of PDK1 and in flies with partial loss of function mutations in two additional insulin signaling genes, Dp110 and Akt. Furthermore, our laboratory previously reported that mutation of chico, which encodes an insulin receptor substrate in Drosophila, also forestalls ARLI (Gargano et al., 2005). All of these data support a role for the insulin signaling pathway in regulating the effects of age on locomotor function in flies.
Disruption of insulin signaling has a number of positive effects on aging in diverse species. An extensive literature demonstrates that loss of function mutations in key genes within the pathway extend life span in worms, flies and mice (Kenyon, 2001; Tatar et al., 2003).
Additionally, extended life span in humans is associated with polymorphisms in the genes that encode the insulin/insulin like growth factor receptor (Suh et al., 2008) and FOXO3A (Willcox et al., 2008), a downstream effector of the pathway. Importantly, though, the effects of the insulin signaling pathway on aging extend beyond its influence on life span. Mutations that disrupt insulin signaling in C. elegans delay senescence of pharyngeal pumping and body movements (Huang et al., 2004) and age-related decline in associative memory (Murakami et al., 2005). Additionally, genetic manipulations that blunt insulin signaling delay cardiac aging in Drosophila (Wessells et al., 2004). The studies reported here and our previous studies on chico mutants (Gargano et al., 2005) demonstrate that the insulin signaling pathway also influences ARLI in flies. Given that the insulin signaling pathway appears to have conserved effects on life span in diverse species, it would be extremely interesting to determine whether this pathway is also involved in the preservation of locomotor function in mammals including humans.
Data from our screen also suggest that genes outside of the insulin signaling pathway might be important for ARLI. Three strains with delayed ARLI contain P-elements inserted between the genes m6 and HLHm7. Additionally, other delayed ARLI strains contain P-element insertions near the genes CG14045, Pfrx and Doc3. While it is tempting to speculate that m6, HLHm7, CG14045, Pfrx or Doc3 might influence ARLI, these effects must be formally validated before definitive connections can be made regarding the role of these genes in locomotor senescence.
Oxidative damage is thought to play a role in ARLI in humans (Nikolic et al., 2005) as well as flies (Bhandari et al., 2007). Consistent with this possibility, many of the transposon insertions with delayed ARLI have elevated resistance to the oxidative stressor paraquat. Several other insertion lines with delayed ARLI, however, have no change in their paraquat survival. These data indicate that while enhanced resistance to paraquat can accompany preservation of locomotor function across age, the enhanced resistance is not required. Additional studies will be necessary to fully delineate the role of oxidative stress resistance in ARLI.
A number of manipulations that extend life span in Drosophila and other species have sex-specific effects (Burger and Promislow, 2004). For example, mutations in Drosophila that blunt insulin signaling produce a larger life span increase in females as compared to males (Tatar et al., 2001; Tu et al., 2002). Additionally, dietary restriction (defined as reduced food intake without malnutrition) causes a greater extension of life span in female versus male flies (Magwere et al., 2004). Interestingly, preservation of locomotor function across age does not strictly follow this same pattern. While the effects of mutations in Dp110 and Akt are more robust in females than in males, the opposite is true for mutations in PDK1. Additional studies will be necessary to formally address whether genetic and other manipulations that preserve locomotor or other functions across age have predictable sex-specific effects in flies and mammals as found for life span.
The authors thank Melissa Borrusch, Pretal Patel, Bill Horton and Heather Wolf for expert technical assistance. The authors thank Laurent Seroude for providing the pGawB collection and for helpful discussions. The authors thank the Bloomington Drosophila Stock Center, Szeged Stock Center and Exelixis Stock Center for providing transposon insertion strains. This work was supported by grants from the National Institute on Aging and the American Federation for Aging Research to M.G.
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.