|Home | About | Journals | Submit | Contact Us | Français|
The C. elegans insulin/IGF-1 signaling (IIS) cascade plays a central role in the regulation of lifespan, dauer diapause, metabolism and stress response. The major regulatory control of IIS is through phosphorylation of its components by serine/threonine-specific protein kinases. In a RNAi screen for serine/threonine protein phosphatases that counter-balance the effect of the kinases in the IIS pathway, we identified pptr-1, a B56 regulatory subunit of the PP2A holoenzyme. Modulation of pptr-1 affects phenotypes associated with the IIS pathway including lifespan, dauer, stress resistance and fat storage. We show that PPTR-1 functions by regulating worm AKT-1 phosphorylation at Thr 350. With striking conservation, mammalian B56β regulates Akt phosphorylation at Thr 308 in 3T3-L1 adipocytes. In C. elegans, this modulation ultimately leads to changes in subcellular localization and transcriptional activity of the forkhead transcription factor DAF-16. This study reveals a conserved role for the B56 regulatory subunit in modulating insulin signaling through AKT dephosphorylation and thereby has widespread implications in cancer and diabetes research.
The insulin/IGF-1-like signaling (IIS) pathway is an evolutionarily conserved neuro-endocrine pathway that regulates metabolism, development, stress resistance and lifespan (Antebi, 2007; Barbieri et al., 2003; Kenyon, 2005; Wolff and Dillin, 2006). In Caenorhabditis elegans (C. elegans), the insulin-like receptor DAF-2 (Kimura et al., 1997) signals through a PI 3-kinase (AGE-1/AAP-1) (Morris et al., 1996; Wolkow et al., 2002) signaling cascade that activates the downstream serine/threonine kinases PDK-1, AKT-1, AKT-2 and SGK-1 (Hertweck et al., 2004; Paradis et al., 1999; Paradis and Ruvkun, 1998). These kinases in turn function to negatively regulate the forkhead transcription factor (FOXO), DAF-16 (Lin et al., 1997; Ogg et al., 1997).
Reduction-of-function mutations in serine/threonine kinases upstream of DAF-16 lead to changes in lifespan, development, metabolism and/or stress resistance (Antebi, 2007; Kenyon, 2005; Wolff and Dillin, 2006). Importantly, loss-of-function mutations in daf-16 completely suppress these phenotypes (Antebi, 2007; Kenyon, 2005; Mukhopadhyay et al., 2006; Wolff and Dillin, 2006). Thus DAF-16 is a major downstream target of the IIS pathway. Regulation of DAF-16 by AKT-1, AKT-2 and SGK-1 results in its nuclear exclusion and sequestration in the cytosol (Lin et al., 2001) (Hertweck et al., 2004; Lee et al., 2001). In contrast, under low signaling conditions, active DAF-16 enters the nucleus and transactivates or represses its direct target genes (Henderson and Johnson, 2001; Hertweck et al., 2004; Lee et al., 2001; Lin et al., 2001; Oh et al., 2006). Strikingly, this negative regulation of FOXO/DAF-16 is conserved across species. In mammals, the Akt and SGK kinases can phosphorylate and negatively regulate FOXO (Brunet et al., 1999; Brunet et al., 2001).
Although regulation of the IIS pathway by serine/threonine protein kinases has been extensively studied, little is known about the phosphatases acting in this pathway. In C. elegans, the lipid phosphatase DAF-18 (homologous to mammalian Phosphatase and Tensin Homolog, PTEN) is the only phosphatase that has been identified and characterized as a negative regulator of the IIS pathway (Gil et al., 1999; Mihaylova et al., 1999; Ogg and Ruvkun, 1998; Rouault et al., 1999). The increased lifespan of daf-2 mutant worms is suppressed by loss-of-function mutation in daf-18 (Dorman et al., 1995; Larsen et al., 1995). Therefore, to identify additional regulators of the IIS pathway, we performed a directed RNAi screen of serine/threonine protein phosphatases that affect phenotypes regulated by the IIS pathway.
C. elegans development proceeds from an egg, through 4 larval stages into a self-fertilizing, hermaphrodite adult. However, under unfavorable growth conditions such as crowding and low food availability, worms enter a stage of diapause known as dauer (Riddle D., 1997). Upon favorable growth conditions, dauers are able to form reproductive adults. Since worms form dauers constitutively when the function of IIS pathway is reduced by mutations, we took advantage of a temperature-sensitive (ts) allele of daf-2 for the RNAi screen (Riddle et al., 1981). We screened for genes that suppressed dauer formation in daf-2(e1370) mutants. In this report, we characterize PPTR-1, a regulatory subunit of the PP2A holoenzyme, as an important regulator of development, longevity, metabolism and stress response in C. elegans. We show that PPTR-1 acts by modulating AKT-1 phosphorylation and as a consequence controls DAF-16 activity.
To identify the serine/threonine phosphatases in the C. elegans genome, we performed in silico analyses using both NCBI KOGs (clusters of euKaryotic Orthologous Groups) and WormBase (a C. elegans database: http://www.wormbase.org; WS152) annotations. A total of 60 genes were identified for further analysis (Figure 1A). We obtained RNAi clones for these phosphatases from the Ahringer RNAi library (Kamath et al., 2003), generated them using available clones from the ORFeome library (Reboul et al., 2003) or cloned them de-novo using Gateway Technology (Invitrogen, USA; Materials and Methods). We were unable to clone 3 of the phosphatase cDNAs and therefore screened a total of 57 candidates.
In addition, we included 6 of the 7 annotated PP2A holoenzyme regulatory subunits (one was not cloned) in the screen for two reasons. First, a preliminary chemical inhibitor screen identified the PP2A family of phosphatases as important regulators of DAF-16 nuclear translocation (Padmanabhan and Tissenbaum, unpublished data). Second, the PP2A holoenzyme is comprised of a catalytic, structural and a regulatory subunit (Janssens et al., 2008) and RNAi of the catalytic and structural subunits of PP2A resulted in lethality (data not shown).
daf-2(e1370) carries a mutation in the insulin receptor tyrosine kinase domain that results in a ts phenotype for dauer formation (Kimura et al., 1997). daf-2(e1370) worms arrest as 100% dauers at 25°C whereas at 15°C they have a normal reproductive cycle (Riddle D., 1997). At an intermediate temperature of 20 °C, a significant percentage of daf-2(e1370) worms form dauers. Therefore, at this temperature, one can use RNAi to easily assess the contribution of any gene in modulating daf-2 dauer formation.
For the screen, daf-2(e1370) mutants were grown on RNAi-expressing bacteria for two generations, and eggs were picked onto 3 plates for each RNAi clone (Figure 1B). The plates were incubated at 20°C and scored 3.5-4 days later for the presence of dauers and non-dauers. Since DAF-18 is the only known phosphatase that negatively regulates the IIS in C. elegans, we used daf-18 RNAi as a positive control in all our experiments. From a total of 63 RNAi clones (57 phosphatases and 6 regulatory subunits), we identified two phosphatases that dramatically decreased daf-2(e1370) dauer formation to a level similar to daf-18 RNAi (Figure 1C).
Our top candidate, fem-2 (T19C3.4) functions in C. elegans sex determination (Kimble et al., 1984; Pilgrim et al., 1995). However, further analysis with an additional daf-2 allele, daf-2(e1368), revealed that fem-2 RNAi suppresses dauer formation in an allele-specific manner. fem-2 RNAi suppressed dauer formation of daf-2(e1370) but not daf-2(e1368) (data not shown) and therefore, we focused on the next top candidate.
pptr-1 (W08G11.4), the next candidate, is a member of the B56 family of genes encoding regulatory subunits of the PP2A protein phosphatase holoenzyme. The C. elegans genome contains 7 known PP2A regulatory subunit genes (pptr-1 and pptr-2, B56 family; sur-6, B55 family; F47B8.3, C06G1.5, rsa-1 and T22D1.5, B72 family; currently F47B8.3 is not annotated as a PP2A regulatory subunit according to WormBase Release WS194). To determine the specificity of pptr-1 in regulating dauer formation, we re-tested the six PP2A regulatory subunits included in the screen for their ability to regulate dauer formation in daf-2(e1370) mutants. Knockdown efficiency of each RNAi clone was verified by RT PCR (Supplemental Table 1A). As shown in Figure 1D and Supplemental Table 1A and 1B, only pptr-1 RNAi suppressed daf-2(e1370) dauer formation comparable to daf-18 RNAi.
We next analyzed the effect of pptr-1 RNAi on dauer formation of daf-2(e1368). pptr-1 RNAi significantly suppressed the dauer formation of daf-2(e1368) (69.2 ± 9.4 % on vector RNAi versus 3.8 ± 4.4% on pptr-1 RNAi; Table 1 and Supplemental Table 2). Therefore the effect of pptr-1 RNAi on daf-2 mutants is not allele-specific. Together these results indicate that pptr-1 may function downstream of daf-2. In addition, pptr-1 is the only PP2A regulatory subunit to affect daf-2 dauer formation.
To further investigate the role of pptr-1 in dauer formation, we performed genetic epistasis analysis. In addition to the C. elegans IIS pathway, a second parallel TGF-β pathway also regulates dauer formation (Patterson and Padgett, 2000; Savage-Dunn, 2005). In this pathway, loss of function mutations in daf-7 (TGF- β ligand), daf-1 and daf-4 (receptors) or daf-14 and daf-8 (R-Smads) lead to constitutive dauer formation; loss-of-function mutations in daf-3 (Co-Smad) or daf-5 (Sno/Ski) suppress these phenotypes (da Graca et al., 2004; Gunther et al., 2000; Inoue and Thomas, 2000; Patterson et al., 1997; Ren et al., 1996). However, mutations in daf-3 do not suppress daf-2(e1370) dauer formation (Vowels and Thomas, 1992). We generated a daf-2(e1370);daf-3(mgDf90) double mutant, which bears a null mutation in daf-3 (Patterson et al., 1997) which essentially removes the input from the TGF- β pathway for dauer formation. In this strain, the dauer formation of daf-2(e1370);daf-3(mgDf90) worms was suppressed by pptr-1 RNAi (94.5 ± 0.8 % dauers on vector RNAi to 42.7 ± 14.6. % dauers on pptr-1 RNAi; Table 1 and Supplemental Table 2). This data suggests that pptr-1 controls dauer formation specifically through the IIS pathway and not the TGF-β pathway.
In addition to dauer formation, the C. elegans IIS pathway also regulates lifespan, fat storage and stress resistance (Antebi, 2007; Kenyon, 2005; Wolff and Dillin, 2006). Since pptr-1 regulates dauer formation specifically via the IIS pathway, we next determined whether this gene could also affect these other important phenotypes.
Mutations in daf-2 result in lifespan extension (Kenyon et al., 1993) that is suppressed by loss-of-function mutations in daf-18 (Dorman et al., 1995; Larsen et al., 1995). To investigate whether pptr-1 can regulate lifespan similar to daf-18, we determined whether knocking down pptr-1 by RNAi could affect daf-2(e1370) lifespan. We grew wild type and daf-2(e1370) worms on vector, daf-18 and pptr-1 RNAi and measured lifespan (Figure 2A). Similar to daf-18 RNAi, knockdown of pptr-1 resulted in a significant reduction in daf-2(e1370) lifespan (mean lifespan of daf-2(e1370) on vector RNAi is 33.9 ± 0.7 days, on pptr-1 RNAi is 27.7 ± 0.9 days and on daf-18 RNAi is 20.4 ± 0.6 days, p value < 0.0001; Figure 2A and Supplemental Table 3A). In contrast, lifespan of wild type was unaffected by pptr-1 RNAi (mean lifespan of wild type on vector RNAi is 22.8 ± 0.4 days, is 21.9 ± 0.5 days on pptr-1 RNAi and 18.6 ± 0.3 days on daf-18 RNAi; Figure 2B and Supplemental Table 3A). Thus, pptr-1 affects lifespan in addition to dauer formation.
Lifespan extension correlates well with increased stress resistance (Lithgow and Walker, 2002; Oh et al., 2005). For example, daf-2(e1370) mutants are not only long-lived but are also extremely resistant to various stresses such as heat and oxidative stress (Honda and Honda, 1999; Lithgow et al., 1995; Munoz and Riddle, 2003). Therefore, we next analyzed the effect of pptr-1 RNAi on the thermotolerance of daf-2(e1370) mutants. As anticipated, pptr-1 RNAi significantly reduced the thermotolerance of daf-2(e1370) mutants (on vector RNAi, daf-2(e1370) had a mean survival of 15.2 ± 0.7 hrs, whereas on pptr-1 RNAi survival was 13.8 ± 0.5 hrs (p value< 0.006). pptr-1 RNAi did not affect the thermotolerance of wild type worms; mean thermotolerance was 9.8 ± 0.4 hrs on vector RNAi, versus 9.3 ± 0.3 hrs on pptr-1 RNAi; Figure 2C and Supplemental Table 4).
In addition to enhanced lifespan and stress resistance, daf-2 mutants have increased fat storage (Ashrafi et al., 2003; Kimura et al., 1997). We next asked whether pptr-1 could also affect fat storage in wild type and daf-2(e1370) worms using Sudan black staining. Consistent with our lifespan and stress resistance results, pptr-1 RNAi suppressed the increased fat storage of daf-2(e1370) without affecting wild type fat storage (Figure 2D). Finally, daf-2 mutants have a slow growth phenotype (Gems et al., 1998; Jensen et al., 2007) that is suppressed by knockdown of daf-16 by RNAi (Supplemental Table 5). Similar to daf-16 RNAi and daf-18 RNAi, pptr-1 RNAi suppresses this slow growth phenotype. Together, these experiments suggest that pptr-1 regulates multiple phenotypes associated with the IIS pathway in C. elegans.
Signals from DAF-2 are transduced to the PI 3-kinase AGE-1 to activate the downstream serine/threonine kinase PDK-1. PDK-1 activates three downstream serine/threonine kinases, AKT-1, AKT-2 and SGK-1 (Antebi, 2007; Kenyon, 2005; Wolff and Dillin, 2006). These kinases together regulate the transcription factor DAF-16 by direct phosphorylation (Hertweck et al., 2004). Mutations in daf-16 suppress the enhanced dauer formation of pdk-1 (Paradis et al., 1999) or akt-1/akt-2 mutants (Oh et al., 2005; Paradis and Ruvkun, 1998). Thus far, our analysis suggests that pptr-1 functions in the IIS pathway. We sought to identify the potential target of pptr-1 by performing genetic epistasis experiments on components of the IIS pathway.
First we analyzed the effect of pptr-1 RNAi on dauer formation of a pdk-1 mutant. Dauer formation of pdk-1(sa680) was suppressed by pptr-1 RNAi (95.6 ± 1.0 % dauers on vector RNAi versus 9.5 ± 0.3 % dauers on pptr-1 RNAi, Table 1 and Supplemental Table 2). In contrast, daf-18 RNAi had no effect on pdk-1(sa680) dauer formation (Table 1 and Supplemental Table 2). Therefore, these results place pptr-1 downstream of pdk-1 and are consistent with the current understanding that daf-18 acts upstream of pdk-1.
Next, to investigate whether pptr-1 acts at the level of akt-1, akt-2 or sgk-1, we first analyzed dauer formation in akt-1(ok525), akt-2(ok393) and sgk-1(ok538) single mutants and the akt-1(ok525);akt-2(ok393) double mutant. While akt-1(ok525), akt-2(ok393) and sgk-1(ok538) single mutants do not arrest as dauers at either 20 or 25 °C, the akt-1(ok525);akt-2(ok393) double mutant forms 100% dauers at all temperatures (Oh et al., 2005). To circumvent this problem, we generated double mutants of daf-2(e1370);akt-1(ok525), daf-2(e1370);akt-2(ok393) and daf-2(e1370);sgk-1(ok538) and tested these strains for dauer formation on vector, daf-18 and pptr-1 RNAi. We reasoned that in a daf-2 mutant background, the akt-1, akt-2 and sgk-1 mutants would exhibit temperature-induced dauer formation. Indeed, all three double mutants were able to form dauers at 20 °C (Table 1 and Supplemental Table 2- see panel for vector RNAi). Importantly, pptr-1 RNAi significantly suppressed dauer formation in daf-2(e1370);akt-2(ok393) (36.8 ± 3.8 % dauers on vector RNAi versus 10.8 ± 4.3 % on pptr-1 RNAi; Table 1 and Supplemental Table 2). In addition, pptr-1 RNAi suppressed dauer formation of daf-2(e1370);sgk-1(ok538) worms (65.4 ± 4.9 % dauers on vector RNAi versus 0 % on pptr-1 RNAi, Table 1 and Supplemental Table 2). In contrast, pptr-1 RNAi did not affect dauer formation of daf-2(e1370);akt-1(ok525) mutants (vector RNAi is 94.8. ± 3.1 % versus 96.0 ± 1.7 % on pptr-1 RNAi; Table 1 and Supplemental Table 2). However, daf-18 RNAi can suppress daf-2(e1370) akt-1(ok525) dauer formation (reduced to 10.5 ± 0.8%; Table 1). These observations genetically place pptr-1 at the level or downstream of akt-1 in the IIS pathway.
Since pptr-1 and akt-1 genetically interact, we wanted to investigate whether they have a common expression pattern. We generated or obtained akt-1gfp, akt-2gfp, sgk-1gfp and we tagged pptr-1 with mCherry and a minimal flag tag to generate pptr-1 mCherry-flag transgenic worms (hence referred to as pptr-1mC-flag; Materials and Methods; GFP/mC-FLAG refers to protein while gfp/mC-flag stands for transgene). We made double transgenic worms by crossing pptr-1mC-flag worms to each of the above-mentioned GFP lines. Similar to published data, we observed AKT-1GFP predominantly in the pharynx, several head neurons, the nerve ring, spermathecae and vulva (Paradis and Ruvkun, 1998); AKT-2GFP in the pharynx (predominantly in the anterior region), somatic muscles, vulva muscles, spermathecae (Paradis and Ruvkun, 1998); SGK-1GFP in amphid neurons, intestine and some tail neurons (Hertweck et al., 2004) (Figure 3A, B, C middle panel and Supplemental Figure 1). PPTR-1mC-FLAG was also observed in the pharynx, head neurons, nerve ring, spermathecae and vulva (Figure 3A, B, C left panel and Supplemental Figure 1). To observe the sub-cellular localization of PPTR-1, we stained pptr-1mC-flag worms with DAPI (Materials and Methods). We find that PPTR-1 is predominantly cytosolic with little DAPI overlap (Supplemental Figure 2 and Supplemental Movies). As shown in Figure 3A-C (Merge) as well as in the confocal movies (Supplemental Movies), there is remarkable overlap between the expression patterns of PPTR-1 and AKT-1. We also observed partial overlap between AKT-2GFP and PPTR-1mC-FLAG, predominantly in the pharynx (Figure 3B, Merge; Supplemental Movies). SGK-1 and PPTR-1 are expressed in different tissues and we do not see any significant overlap (Figure 3C, Merge; Supplemental Movies).
Given the genetic epistasis as well as the overlapping expression patterns, we next determined whether PPTR-1 directly interacts with AKT-1 by co-immunoprecipitation (co-IP) in C. elegans. For all biochemical experiments, we used the PD4251 strain as a control. This strain contains Pmyo-3gfp with a mitochondrial localization signal and Pmyo-3lacZ-gfp with a nuclear localization signal (Fire et al., 1998). This strain will be referred to as myo-3gfp. We prepared lysates from mixed-stage akt-1gfp; pptr-1mC-flag and myo-3gfp; pptr-1mcherry-flag transgenic worms. Following immunoprecipitation with either anti-FLAG or anti-GFP antibody, we found that PPTR-1 specifically interacts with AKT-1 and not with MYO-3GFP (Figure 4A; Materials and Methods). We also performed co-IP experiments to investigate whether PPTR-1 and AKT-2 interact, since we observed partial overlap in expression pattern of these proteins. We find that PPTR-1 does not interact with AKT-2 (Supplemental Figure 3). Our epistasis analyses show no genetic interaction between pptr-1 and sgk-1. Moreover, we observe no overlap in the expression pattern of these two proteins using confocal microscopy (Supplemental Movies). However, we find that PPTR-1mC-FLAG and SGK-1GFP can interact in our co-IP experiments (Supplemental Figure 3). We do not believe this biochemical interaction to have a measurable functional output and did not pursue it further.
In mammals, Akt is activated by PDK phosphorylation at Thr 308 and PDK-2/TORC-2 protein complex at Ser 473 (Brazil and Hemmings, 2001; Sarbassov et al., 2005). In C. elegans AKT-1, these sites correspond to Thr 350 and Ser 517, respectively. We generated affinity-purified phospho-specific antibodies (21st Century BioChemicals, USA; Materials and Methods) against both sites to further investigate the role of PPTR-1 on AKT-1 phosphorylation. Following immunoprecipitation with anti-GFP antibody from either akt-1gfp or akt-1gfp;pptr-1mC-flag strain, we compared the phosphorylation status at these two sites. We find that overexpressing PPTR-1 can dramatically decrease the phosphorylation of the T350 site while having a marginal effect on the Ser 517 site (Figure 4B). As a control experiment, we treated the immunoprecipitated AKT-1GFP samples with lambda phosphatase and observed loss of the Thr and Ser phosphorylation, showing the specificity of the phospho-AKT antibodies (Supplemental Figure 4A). Thus, in C. elegans, PPTR-1 functions by directly regulating the dephosphorylation of AKT-1 primarily at the Thr 350 (mammalian Thr 308) site.
Given the evolutionary conservation of the C. elegans IIS pathway, we next determined if this mechanism of AKT-1 dephosphorylation mediated by PPTR-1 is also conserved in mammals. The mammalian B56 family of PP2A regulatory subunits has 8 members encoded by 5 genes that express in different tissues (Eichhorn et al., 2008). We used 3T3-L1 adipocytes to perform these studies since in this system, there is a well-characterized insulin signaling pathway that is responsive to changes in insulin levels (Ugi et al., 2004; Watson et al., 2004). We first compared microarray data from the expression profiles of fibroblasts to differentiated 3T3-L1 adipocytes (Powelka et al., 2006) to determine which B56 members were expressed in the adipocytes. We identified 2 genes, PPP2R5A (B56α) and PPP2R5B (B56β) as the top candidates. We knocked down either one or both these regulatory subunits by designing Smartpool siRNAs (Dharmacon, USA) and verified the silencing by quantitative RT PCR (Supplemental Figure 3B). Serum-starved siRNA-treated 3T3-L1 adipocytes were then stimulated with increasing concentrations of insulin. The cells were lysed and the proteins analyzed by western blotting using mammalian Akt phospho-specific antibodies (Materials and Methods). Knockdown of B56β results in a dramatic increase in phosphorylation at the Thr 308 site of Akt with relatively less changes in Ser 473 phosphorylation (Figure 4C). However, silencing of B56α had no effect on the phosphorylation status of Akt at either site. We observed that siRNA against both the PP2A catalytic subunits (PP2Acα/β) results in increased phosphorylation at Thr 308 but not at Ser 473. Together, our data suggests that PPTR-1/B56β regulatory subunits function to modulate AKT-1 phosphorylation in a conserved manner across phylogeny.
We next determined the consequences of modulating PPTR-1 dosage on the IIS pathway. In C. elegans, one of the major targets of AKT-1 is the forkhead transcription factor, DAF-16. Active signaling through the IIS pathway results in the phosphorylation of DAF-16 by AKT-1, AKT-2 and SGK-1, leading to its nuclear exclusion (Antebi, 2007; Kenyon, 2005; Wolff and Dillin, 2006). However, under low signaling conditions, DAF-16 translocates into the nucleus, where it can directly bind and activate/repress the transcription of target genes involved in dauer formation, lifespan, stress resistance and fat storage (Oh et al., 2006). We asked whether pptr-1 regulates IIS pathway-specific phenotypes by modulating DAF-16 function. Since we observed reduced phosphorylation of AKT-1 upon overexpression of PPTR-1, we first looked at the effect of PPTR-1 overexpression on DAF-16 nuclear localization (Henderson and Johnson, 2001; Lee et al., 2001; Lin et al., 2001). We generated a daf-16gfp;pptr-1mC-flag strain and then compared the DAF-16 nuclear localization in these worms with a daf-16gfp strain (Figure 5A and Supplemental Figure 5A). We categorized DAF-16GFP localization as completely cytosolic, mostly cytosolic, mostly nuclear or completely nuclear. We find that DAF-16GFP nuclear localization is enhanced when PPTR-1 is overexpressed (Figure 5A and Supplemental Figure 5A). To determine the specificity of this response, we used mCherry RNAi to effectively knockdown mCherry expression in pptr-1mC-flag thereby reducing the expression of pptr-1 transgene (data not shown). Our results show that the enhanced nuclear localization upon PPTR-1 overexpression is suppressed when pptr-1mC-flag;daf-16gfp worms are grown on mCherry RNAi (Figure 5A and Supplemental Figure 5A) and mCherry RNAi has little effect on DAF-16 localization in daf-16gfp worms. These experiments suggest that increased dosage of pptr-1 affects DAF-16 nuclear localization. Consistent with its role in the C. elegans IIS pathway, we find that overexpression of pptr-1 significantly increases the lifespan of wild type worms but does not further enhance the lifespan daf-2(e1370) worms (Figure 5B and Supplemental Table 3B; mean lifespan of wild type is 23.9 ± 0.3 days, pptr-1mC-flag is 30.1 ± 0.5 days, p<.0001, and the unc-119(+); unc-119(ed3) control strain is 22.6 ± 0.3 days).
As a corollary to this experiment, we next looked at the effect of pptr-1RNAi on DAF-16 nuclear localization. For this, we generated a strain with a daf-2(e1370);daf-16gfp strain. At the permissive temperature of 15°C, DAF-16GFP is excluded from the nucleus in the daf-2(e1370);daf-16gfp strain. However, at the non-permissive temperature of 25 °C, progressive nuclear localization of DAF-16GFP is observed. We grew daf-2(e1370);daf-16gfp worms on either vector, pptr-1 or daf-18 RNAi and measured the extent of nuclear localization at 25 °C. We find that pptr-1 RNAi significantly reduced DAF-16 nuclear localization, similar to the effect of daf-18 RNAi (Figure 5C and Supplemental Figure 5B). Together, these experiments suggest that changes in PPTR-1 levels affect the activity of AKT-1 and as a result, modulate DAF-16 sub-cellular localization.
DAF-16 regulates the transcription of many downstream genes such as sod-3, hsp-12.6, sip-1 and mtl-1 (Furuyama et al., 2000; Lee et al., 2003; McElwee et al., 2003; Murphy et al., 2003; Oh et al., 2006). We next tested the effects of pptr-1 RNAi on these DAF-16 transcriptional targets. We first tested sod-3 which has been shown to be a direct target of DAF-16 by chromatin immunoprecipitation (Oh et al., 2006) and its expression changes in response to modulation of the IIS pathway (Furuyama et al., 2000; Libina et al., 2003; Murphy et al., 2003). We grew a daf-2(e1370);Psod-3gfp(muIs84) strain on either vector, daf-18 or pptr-1 RNAi to look at the effect on GFP expression. Similar to worms grown on daf-18 RNAi, pptr-1 RNAi reduces expression of GFP (Figure 5D and Supplemental Figure 5C). Therefore, modulation in the levels of pptr-1 can affect the expression of direct DAF-16 target genes.
We further analyzed the expression of known DAF-16 target genes by quantitative RT-PCR in a daf-2(e1370) mutant background. As a control, we analyzed whether each of these target genes expressed in a daf-16-dependent manner as previously reported (McElwee et al., 2003; Murphy et al., 2003; Oh et al., 2006). As shown in Figure 5E and Supplemental Figure 5D, daf-16 RNAi dramatically suppressed the expression levels of these genes. Next, we tested the effects of either pptr-1 or daf-18 RNAi on the expression of these genes. We found that pptr-1 RNAi also suppressed the expression of these genes to a level similar to daf-18 RNAi. Taken together, our data suggests that PPTR-1 positively regulates DAF-16 nuclear localization and thereby its activity.
In a directed RNAi screen for Ser/Thr phosphatases in the C. elegans IIS pathway, we identified pptr-1, a B6 regulatory subunit of the PP2A holoenzyme. PP2A itself is a ubiquitously expressed phosphatase that is involved in multiple cellular processes including regulation of insulin signaling by direct dephosphorylation of Akt (Andjelkovic et al., 1996; Resjo et al., 2002; Ugi et al., 2004). Substrate specificity of PP2A is achieved by its association with distinct regulatory subunits. Our studies provide a mechanistic insight into how the C. elegans PP2A regulatory subunit PPTR-1 modulates insulin signaling by specifically regulating AKT-1 phosphorylation and activity in the context of a whole organism.
In our model (Figure 5F), PPTR-1 acts to negatively regulate signals transduced through the IIS pathway, ultimately controlling the activity of the FOXO transcription factor DAF-16. Under low signaling conditions, DAF-16 is able to translocate to the nucleus and transactivate or repress its downstream targets. It is well established that AKT modulates DAF-16 sub-cellular localization. Thus, the activity of AKT-1, as governed by its phosphorylation status, directly translates into the activity of DAF-16. In this study, we show that PPTR-1 directly interacts with AKT-1 and regulates its activity by modulating its phosphorylation, predominantly at the Thr 350 site. Less active AKT-1 results in increased DAF-16 nuclear localization. Indeed, DAF-16 is found to be more nuclear throughout the worm when PPTR-1 is overexpressed. As a corollary, knocking down pptr-1 by RNAi results in less nuclear DAF-16 as well as reduced expression of DAF-16 target genes such as sod-3, hsp-12.6, mtl-1 and sip-1. These genes are known to play a combinatorial role in adaptation to various stresses, leading to enhanced dauer formation and increased lifespan. Consistent with the decreased levels of these important genes, pptr-1 RNAi results in a significant decrease in the dauer formation, lifespan, thermotolerance of daf-2(e1370) mutants. In addition, pptr-1 also regulates other DAF-16-dependent outputs of the IIS pathway such as fat storage. Thus, we find that normal levels of pptr-1 are important under low insulin signaling conditions. However, pptr-1 RNAi does not affect IIS pathway-associated phenotypes in wild type worms. There could be several reasons for this observation. Firstly, under normal signaling conditions, AKT-1, AKT-2 as well as SGK-1 are active and negatively regulate DAF-16. Therefore, changes in the AKT-1 activity alone brought about by pptr-1 RNAi may not have a significant effect on DAF-16-dependent phenotypes. Secondly, PPTR-1 itself may be negatively regulated by the IIS pathway, leading to increased AKT-1 phosphorylation. Along similar lines, in mammals, insulin signaling can downregulate the expression and activity of the PP2A catalytic subunit (Hojlund et al., 2002; Srinivasan and Begum, 1994; Ugi et al., 2004). Thus, under normal conditions, further down regulation of pptr-1 by RNAi may have no effect. We speculate that in C elegans, in response to changing environmental cues, PPTR-1 helps to downregulate the insulin signaling pathway to promote DAF-16 activity, enabling the worm to either enter diapause or enhance its tolerance to stress as adults.
In mammals, Akt controls a myriad of secondary signaling cascades that regulate glucose transport, protein synthesis, genomic stability, cell survival and gene expression (Toker and Yoeli-Lerner, 2006). Previous studies have implicated roles for PP2A and PHLPP phosphatases in the negative regulation of Akt (Kuo et al., 2008). The PP2A inhibitor Okadaic acid can increase Akt phosphorylation predominantly at Thr 308 and enhance glucose transport in adipocytes (Rondinone et al., 1999). Consistent with this, our results show that siRNA knockdown of the PP2A catalytic subunit and more importantly, the B56β regulatory subunit results in enhanced Akt phosphorylation at Thr 308 in 3T3-L1 adipocytes. Thus, our study points at the remarkable functional conservation of the B56/PPTR-1 regulatory subunit of PP2A in regulating AKT phosphorylation between C. elegans and higher mammals. In worms, we also see a modest effect on Ser 517 (equivalent to mammalian Ser 473) phosphorylation by PPTR-1 overexpression. However, we do not observe a difference in Ser 473 phosphorylation in adipocytes. This difference may be explained by the fact that in worms, we are determining the phosphorylation of AKT-1 in the context of a whole organism. Additionally, in mammals phosphorylation state of one Akt site may influence the status of the other (Gao et al., 2005; Toker and Newton, 2000). We do not see a role for the PP2A B55 subunit (sur-6) in the C. elegans IIS pathway. However, a recent report using cell culture has implicated the mammalian B55 in the regulation of AKT (Kuo et al., 2008).
Dysregulation of Akt has been implicated in diseases such as cancer and diabetes (Rondinone et al., 1999; Sasaoka et al., 2006; Smith et al., 1999; Zdychova and Komers, 2005). In fact, the onset of diabetes is often associated with changes in Akt phosphorylation (Zdychova and Komers, 2005). In several cancer models, loss of function mutations in the PTEN results in hyper-phosphorylated and activated Akt (Groszer et al., 2001; Hakem and Mak, 2001; Stiles et al., 2002; Testa and Bellacosa, 2001) Our studies show that like PTEN, PPTR-1 acts to negatively regulate the insulin/IGF-1 signaling. Given the important role of PPTR-1/B56 in modulating Akt activity, this protein may be a potential therapeutic target for the treatment of diabetes as well as cancer.
All strains were maintained at 15°C using standard C. elegans techniques (Stiernagle, 2006). Double mutants were made using standard genetic methods while transgenic worms were made by microparticle bombardment as described in the Supplemental Materials and Methods. For all RNAi assays, the worms were grown for at least two full generations on the RNAi bacteria. Preparation of RNAi plates is described in the Supplemental Materials and Methods.
C. elegans assays were modified from previously published methods (Henderson and Johnson, 2001; Kimura et al., 1997; Libina et al., 2003; Oh et al., 2006; Oh et al., 2005). For detailed description of these assays see Supplemental Materials and Methods.
Transgenic worms were grown in three 100 mm plates seeded with OP50 bacteria at 20 °C. Worms were harvested by washing with M9 buffer and pellet collected by centrifugation. The pellet was resuspended in 250 μl lysis buffer (20 mM Tris-Cl, 137 mM NaCl, 10% glycerol, 1% Triton X-100, 25 mM β-glycerophosphate, Protease inhibitor cocktail (Roche Biochemicals, USA), pH 7.4). The worms were sonicated with Bioruptor (Diagenode, USA) using maximum power output (1 min sonication, 2 min off-repeated 10 times). The lysate was cleared by centrifugation and protein content estimated by Bradford method. Lysate equivalent to 1.5 mg total protein was pre-cleared with 50 μl of protein-G agarose beads, fast flow (Upstate, USA) and then immunoprecipitated overnight at 4 °C using either anti-GFP monoclonal antibody (Sigma, USA) or anti-FLAG M2 gel (Sigma, USA). The following morning, 50 μl protein-G agarose beads, fast flow were added to the GFP IP to capture the immune complex. The agarose beads were then washed 5 times with lysis buffer. Following this step, the beads were boiled in Laemelli's buffer.
For western blot analysis, immunoprecipitated protein samples was resolved on a 10% SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blocked in TBST (Tris Buffered Saline containing 0.05% Tween 20, ph 7.4) containing 5% non-fat milk at RT for 1 hour. Membranes were then washed with TBST and incubated overnight with 1:1000 dilutions of antibodies in TBST containing 5% non-fat milk 4 °C. Membranes were washed 3 times with TBST and then incubated with TBST containing 5% non-fat milk containing a 1:10,000 dilution of the secondary antibody. Antibodies used for western were:
Transgenic worms were grown at 20 °C in 3-4 large (100 mm) plates seeded with OP50. Worms were collected by washing with 1 X PBS and the pellet was then immediately frozen in dry ice. Around 500 μl lysis buffer, supplemented by Sigma Phosphatase inhibitor cocktails I and II (50×) and Protease inhibitor cocktail (Roche Biochemicals, USA) was added to the pellet and sonicated using a Misonix (3000) sonicator (Misonix, USA; power output set at 4, 3 pulses of 10 secs each with 1 min interval between pulses). The lysates were clarified by centrifugation at 13000 rpm for 10 mins at 4°C and the protein content estimated by Quick Bradford (Pierce). About 3.5 μg of anti-GFP monoclonal antibody (3E6, Invitrogen USA) was used for each IP from lysates containing 1.7 mg protein in a volume of 1ml. IPs were performed overnight at 4°C and antibody-protein complexes were captured using 50 μl of protein-G agarose beads, fast flow (Upstate, USA) for 2 hrs at 4°C. The pellets were washed 3 times with lysis buffer supplemented by protease and phosphatase inhibitors and boiled in Laemelli's buffer. The IP samples were then resolved on a 10% SDS-PAGE, western blotted and analyzed with phospho-specific antibodies (Supplemental Materials and Methods).
3T3-L1 adipocytes were cultured and differentiated as previously described (Tesz et al., 2007). For siRNA transfections, cells from 4 days post-induction of adipocyte differentiation were used as previously described (Tang et al., 2006). Cells were stimulated with increasing concentrations of insulin and isolated proteins were analyzed by Western blotting Detailed procedure is described in Supplemental Materials and Methods.
Total RNA was isolated using Trizol (Invitrogen, USA) and real-time PCR was performed according to manufacturer's instructions (Applied Biosystems, USA) as detailed in Supplemental Materials and Methods.
We are grateful to Eun-soo Kwon, Kelvin Yen, Craig Mello and Marian Walhout for advice and critical comments on the manuscript, Paul Furcinitti of the UMass Medical School Digital Light Microscopy Core Facility for help with the confocal microscopy, Nina Bhabhalia for technical support and Sandhya Pande for help with the phospho-westerns. We thank Marian Walhout, Gary Ruvkun, Maren Hertweck and Ralf Baumeister for plasmids and strains. Some of the strains were kindly provided by Theresa Stiernagle at the Caenorhabditis Genetics Center, which is funded by the National Institutes of Health National Center for Research Resources. H.A.T. is a William Randolph Hearst Young Investigator. This project was funded in part by NIA (AG025891), the Glenn Foundation an endowment from the William Randolph Hearst Foundation.
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.