Knockdown of Indy/CeNac2 extends Caenorhabditis elegans life span by inducing AMPK/aak-2

Effect of Indy/CeNAC2 knockdown on life span

Knockdown of Indy/CeNAC2 by feeding wildtype N2 worms with bacteria expressing Indy/CeNAC2-specific siRNA caused a significant increase in average life span of the organism (Fig. 1A). Individual data from 4 independent experiments are given in Table 1a-d. Mean life span of the worms fed on bacteria harboring Indy/CeNAC2-specific siRNA was 19.3 ± 0.52 days; mean life span of the worms fed on bacteria harboring the empty vector L4400 was 15.8 ± 0.29 days (Tab.1b). The increase in average life span induced by Indy/CeNAC2 knockdown was +22% (Log Rank p ≤ 0.001). A partial loss of function daf-2 mutant strain (CB1370) was included as a positive control in the experiments and showed the documented life span extension (Table 1a-d) [18]. Thus, our studies confirm a life span extending effect of reduced Indy/CeNAC2 expression in C.elegans.

Figure 1. (A) Representative survival curves for N2 (WT), fed with E. coli GC363 harboring the empty vector L4400 (empty vector control) versus E. coli GC363 harboring INDY/CeNAC2-specific siRNA. INDY/CeNAC2-leads to a significant life span extension (P < 0.001). Life span experiments were repeated four times in independent experiments, figure shows the second experiment, other data in Table 1.(B) Oil Red O staining of N2 wild type fed with E.coli GC363 harboring the empty vector L4400 (empty vector control) versus E.coli GC363 harboring INDY/CeNAC2-specific siRNA (C). N2 wild type worms showed more intensive staining and higher total area stained compared to RNAi treatment (D), Densitometric analysis of B and C, **p = 0.005. (E) Immunoblot analysis of different C. elegans strains (N2, TG) and fasted mouse liver lysate (C) as positive control for phosphorylation (activation) of AMPK/ aak-2 after knockdown of INDY/CeNAC2 (N2/I). Down regulation of INDY/CeNAC2 leads to activation of AMPK / aak-2 compared to control fed wildtype N2. Negative control TG38 (TG) with missing phosphorylation site for AMPK-Thr172 shows no activation when fed with empty vector control neither when treated with RNAi against INDY/CeNAC2 (G/i). Control (c) = liver lysates from fasted mice.

Mean life span (SEM) of worms fed with E.coli GC363 harboring the empty vector L4400 (empty vector control) versus E.coli GC363 harboring INDY/CeNAC2-specific siRNA and CB1370 daf-2(e1370) fed with empty vector control. N2 wild type worms were used as control. Strain TG38 and CB1370 are short and long lived controls, respectively. n = number of worms used in the experiment. Mean life span and SEM are in days.

Effect of Indy/CeNAC2 knockdown on body fat content

In order to test whether or not reduction of Indy/CeNAC2 in C.elegans also shares characteristics of caloric restriction, whole body fat stores were analyzed by phase contrast microscopy using the azo compound Oil Red O to stain triglycerides. The intensity (area by pixel) of Oil Red O staining of worms fed on bacteria harboring Indy/CeNAC2-specific siRNA was ~ 50% lower compared to control worms that were fed on bacteria harboring empty vector L4400 (Indy/CeNAC2-specific siRNA: 5.4e+006 ± 76988, control L4400: 1.016e+007 ± 1.127e+006, p = 0.006 Wilcoxon test) (Fig. 1B-D).

Knockdown of CeNAC2 leads to the activation of AMPK

Given the life span extending effect of Indy/CeNAC2 knockdown together with a reduction in body fat content, we were interested in determining a unifying molecular mechanism. In mice, deletion of mIndy reduced the hepatocellular ATP/ADP ratio and activated hepatic AMPK [6, 19]. The C. elegans AMPK homolog aak2, has been shown to mediate the life span extending effect of various short-term stressors [17] including glucose restriction [20]. Phosphorylation of AMPK/aak2 at threonine 172 is widely used as a surrogate marker of activity of this kinase. Therefore, the phosphorylation of aak2 in worms fed on bacteria harboring Indy/CeNAC2 siRNA was analyzed. Western blot analysis of cell extracts was carried out using an antibody against phospho-Thr¹⁷² of the human AMPK. AMPK/aak2 was markedly phosphorylated in worms with reduced Indy/CeNAC2. A lower state of phosphorylation of AMPK/aak2 was observed for wildtype N2 under control conditions, fed with bacteria harboring the empty vector L4400. Blotted extracts of worms with a deletion in the aak2 gene (TG38 aak-2(gt33)), including the phosphorylation site, showed no signal for phosphorylated AMPK/aak2 (Fig. 1E).

Effect of glucose restriction on life span in Indy/CeNAC2 knockdown

AMPK/aak2 mediates the life span extending effect of glucose restriction in C. elegans [20]. Since AMPK/aak2 was activated in Indy/CeNAC2 RNAi treated worms, and Indy/ceNAC2 encodes a transporter for citrate, a metabolite interfering with glycolysis [11, 21], altered rates of substrate flux through glycolysis might be a unifying principle to couple life span extension via AMPK to the knockdown of Indy/CeNAC2. Glycolysis can be inhibited by 2-Deoxy-D-glucose (DOG), a molecule which competitively inhibits the generation of glucose-6-phosphate from glucose at the phosphoglucokinase level. This is due to the 2-hydroxyl group replaced by hydrogen, rendering DOG incompatible for further glycolysis, while metabolism of other nutrients remains unaffected. Therefore, we tested whether or not DOG administration is still effective in increasing life span in Indy/ceNAC2 knockdown C. elegans. Exposure of 5 mM DOG to the nematodes caused a significant increase in average life span of the organism by 20 ± 1.3% (Log Rank p ≤ 0.001). Exposure of DOG to nematodes with simultaneous knockdown of the Indy homolog CeNAC2 did not show any further life span extension (Log Rank p = 0.135) (Fig. 2A, Tab.2a-c), suggesting that both interventions converge into well related downstream effect.

Figure 2. (A) Survival curves for TG38 AMPK/akk-2(gt33), fed with E. coli GC363 harboring the empty vector L4400 (empty vector control) versus E. coli GC363 harboring INDY/CeNAC2-specific siRNA and CB1370 daf-2(e1370) fed with empty vector control. (B) Survival curves for N2 (WT), fed with E.coli GC363 harboring the empty vector L4400 (empty vector control) versus E.coli GC363 harboring INDY/CeNAC2 siRNA with 2-Deoxy-D-Glucose (DOG+) in the media.

Role of AMPK/aak2 in the life span extending effect of Indy/CeNAC2

In order to determine whether or not the life span extending effect of the reduction of Indy/CeNAC2 depends on AMPK/aak2, life span studies were repeated in C. elegans with an inactivating mutation of AMPK/aak2, TG38 aak-2(gt33). In these TG38 aak-2(gt33) worms, RNAi-mediated knock down of Indy/CeNAC2 did not increase average life span, while we continued to observe the life span extending effect in the N2 wild type (Fig. 2A and Tab.1a-d). Mean life span of TG38 aak-2(gt33)/empty vector control did not differ from TG38 aak-2(gt33)/ Indy/CeNAC2 in all experiments. Long-lived CB1370 daf-2 (e1370) worms, a positive control for life span extension, showed the expected increase in life span in all experiments (Tab.1a-d). To confirm that the mutation in AMPK/aak2 did not by itself alter the expression of Indy/CeNAC2, expression levels were determined by qPCR. Expression levels of Indy/CeNAC2 in TG38 (aak-2(gt33)) Indy/CeNAC2 siRNA treated worms were lowered by 58% (p = 0.05) compared to TG38 aak-2(gt33)/empty vector control treated worms.

Strains and culture conditions

C. elegans strains Wild-type N2, TG38 aak-2(gt33) and E.coli OP50 and GC363 were obtained from the Caenorhabditis Genetics Center (Minneapolis, MN). The gt33 allele has a 606-bp deletion. Wild-type N2 and aak-2 mutant worms were maintained at 22°C on Nematode Growth Medium (NGM) plates seeded with E. coli OP50 using standard methods [32] except for experiments where other is specified.

Bacterially-mediated RNAi

Genetically modified bacteria strain (E. coli HT115 ) for CeINDY was retrieved from Source BioScience. Carrying the L4400 vector, which contains convergent T7 polymerase promoters separated by a multicloning site. E. coli GC363 carry the empty L4400 vector and are considered as the control. Control and siRNA expressing E.coli strains were grown over night on LB Agar at 37°C containing 100 μg/ml ampicillin and 12.5 μg tetracyclin. A single colony was picket and transferred into LB-Media and cultivated over night for a starter culture. Preceding every experiment for bacterial-mediated RNAi, a fresh culture was grown for 6 h in LB-Media from an aliquot of the starter culture. 4–7 days ahead of every experiment the according NGM plates were prepared containing 100 μg/ml ampicillin (Roth) and 1 mM Isopropyl-β-D-thiogalactopyranoside (IPTG, Sigma) as well as 0.1 mg/ml 5-fluoro-2′-deoxyuridine (FUDR, Sigma) if needed.

Life span assays

Worms were cultured at 20°C on Nematode Growth Medium (NGM) plates seeded with E. coli OP50 using standard methods. Prior to the experiments worms were synchronized via 20% Alkaline Hypochlorite Solution using standard protocol [32], eggs were transferred to new plates seeded with OP50. At the L4 (t = 0) moult, animals were transferred to plates containing ampicillin, FUDR and IPTG. RNAi expressing bacteria, were received as sequence verified strains by BioSource (Nottingham, GB). RNAi mediated knockdown of CeNAC2 in N2 was conducted as described by Fei et al. 2003 with significant down regulation of gene expression. As mentioned above seeded with RNAi expressing bacteria or control bacteria (empty vector control). Animals were transferred to new plates every second day and counted. Animals were judged dead when they did not move after repeated prodding with a pick. Animals that crawled off the plates were not included in the analysis (Kaplan Meyer). Assays were repeated independently four times for validation. Plates for life span assay with DOG were prepared from the same bath of NGM agar. The respective chemical was added to a final concentration of 5 mM from a sterile 0.5 M stock solution for DOG. DOG treatment lasted for seven days after larvae reached the L4 moult. Assay was conducted until all worms were either censored or counted as dead as described above.

Expression

Parallel to life span assays nematodes were cultured on plates containing ampicillin, FUDR and IPTG as mentioned above for 5 days, regularly fed with siRNA expressing bacteria. After 5 days worms were washed off plates with M9-Buffer and shortly centrifuged. Supernatant was removed and 500 μl TRIZOL (Life Technologies) was added to the worm pellet. For lysis of worms a Speed Mill (Analytik Jena) was used. Samples were centrifuged for 5 min at full speed, supernatant was transferred into a new tube and 100 μl was added per 500 μl of supernatant volume. Followed by an incubation for 2 min on ice, the samples were centrifuged for 15min at 4°C (12000xg). The polar phase was removed and the nucleic acids were precipitated via isopropyl and washed with ethanol (70%). RNA was resuspended in nuclease free water. Residues of DNA were digested using DNase I, Rnase-free (Thermo Scientific), for cDNA synthesis RevertAid First Strand cDNA Synthesis Kit was used. For detection SYBR Green and 7300 Real Time RCE System by Apllied Biosystems was used. Once knockdown was established it was checked periodically.

Western blot analysis

Synchronized Wild-type N2 and aak-2(gt33) eggs were transferred to NGM plates with GC363 as the main food source. At 22°C eggs were allowed to hatch und larvae to grow until L4 stage. At the first adult day worms were transferred to NGM-plates containing IPTG seeded with siRNAi producing bacteria. After 4 days of feeding, worms were washed from NGM plate using M9 buffer. Worms were set on ice to assemble on the bottom of the tube. After removing the supernatant, the worms were transferred to a tube containing beeds (Carl Roth GmbH) for lysis (Speed Mill Plus, Analytik Jena) and 2 volumes RIPA buffer (150 mM NaCl, 10 nM Tris, pH 7.2, 1% Triton X-100, 1% Desoxycholate, 5 mM EDTA) was added to the pellet. After lysis samples were centrifuged at maximum speed for 5 min (4°C) and supernatant was transferred into a fresh tube. Protein concentration was determined via BCA-Assay (Thermo Scientific). 10 μg of protein from starved mouse liver lysate (C) as positive control was used along with 30 μg of protein from each lysed worm sample for electrophoresis. Samples were electrophoresed on an 10% SDS-polyacrylamide gel and electroblotted onto a PVDF membrane (Milipore). The membrane was incubated in a blocking solution containing anti-phospho-AMPK/ anti-AMPK antibody (1:1000dilution; #2535/ #2532, Cell Signaling Technology) or anti-tubulin mouse antibody (1:1000 dilution; #ab7291, Abcam) at 4°C overnight, followed by incubation with anti-mouse or anti-rabbit horseradish peroxidase antibody (1:5000 dilution; #7076/ #7074S Cell Signaling Technology) for 2 h at 25°C. Stripping Reagent by Roche was used. The ECL Western blotting analysis system (Roche) was used to detect the secondary antibodies on the membrane. Luminescence of the blot was captured using Chemidoc imaging system (Bioad).

Oil red O staining

Worms were cultured at 20°C on Nematode Growth Medium (NGM) using standard methods. Prior to the experiments worms were synchronized via 20% Alkaline Hypochlorite Solution using standard protocol [32], eggs were transferred to new plates for hatching as previously described for life span assays. At the L4 (t = 0) moult, animals were transferred to plates containing ampicillin, FUDR and IPTG and stained with Oil Red O after 7 days. For Oil Red O staining, animals were washed in M9 buffer, and fixed with 4% paraformaldehyde for 15 minutes, were washed again and stained in filtered Oil Red O staining solution (60% Oil Red O stock solution (5 mg/ml isopropanol)/ 40% distilled water) for 30 minutes. After washing in M9, worms were analyzed using BZ-II Analyzer (KEYENCE, Germany) microscope. Photographs of 8–12 worms per condition were taken and at 40x magnification at a fixed exposure time.

Statistical analysis

Statistical analyses for Indy/CeNAC2 knock down and fat content were performed by Wilcoxon test. For comparing significant distributions between different groups in the life span assays statistical calculations were carried out using the log-rank rest (Kaplan Meier) as well as the χ²-Test. Given data shows mean values ±SEM if not described otherwise. All calculations were performed using Microsoft Office Excel 2010 and SPSS version 20.0. Experiments were performed in triplicate except where stated otherwise. Error measures were calculated form the variability of the independently prepared biological samples of the same treatment group where applicable. Differences were considered statistically significant at p < 0.05, therey indicating a probability of error lower than 5%.