REDD2 is enriched in skeletal muscle and inhibits mTOR signaling in response to leucine and stretch

Abstract

The protein kinase mammalian target of rapamycin (mTOR) is well established as a key regulator of skeletal muscle size. In this study, we determined that the stress responsive gene REDD2 (regulated in development and DNA damage responses 2) is a negative regulator of mTOR signaling and is expressed predominantly in skeletal muscle. Overexpression of REDD2 in muscle cells significantly inhibited basal mTOR signaling and diminished the response of mTOR to leucine addition or mechanical stretch. The inhibitory function of REDD2 on mTOR signaling seems to be mediated downstream or independent of Akt signaling and upstream of Rheb (Ras homolog enriched in brain). Knock down of tuberous sclerosis complex 2 (TSC2) using small interfering (si)RNA potently activated mTOR signaling and was sufficient to rescue REDD2 inhibition of mTOR activity, suggesting that REDD2 functions by modulating TSC2 function. Immunoprecipitation assays demonstrated that REDD2 does not directly interact with either TSC1 or TSC2. However, we found that REDD2 forms a complex with 14-3-3 protein and that increasing expression of REDD2 acts to competitively dissociate TSC2 from 14-3-3 and inhibits mTOR signaling. These findings demonstrate that REDD2 is a skeletal muscle specific inhibitory modulator of mTOR signaling and identify TSC2 and 14-3-3 as key molecular links between REDD2 and mTOR function.

Fig. 1.

Expression of REDD1 (regulated in development and DNA damage responses 1) and REDD2 mRNA in mouse tissues. Relative expression levels of REDD1 and REDD2 mRNA were determined by semiquantitative RT-PCR from skeletal muscle (P, plantaris; TA, tibialis anterior; D, diaphragm) and non-skeletal muscle tissues (B, brain; H, heart; Lu, lung; Li, liver; K, kidney) in mouse. RPL26 mRNA expression was used as internal control.

Fig. 2.

REDD2 inhibits leucine- and mechanical strain-activated mTOR signaling. C2C12 myoblasts were cotransfected with S6K1-glutathione S-transferase (GST) or GST-Akt and empty or hemagglutinin (HA)-tagged REDD2. To determine the effects of REDD2 overexpression on S6K1 and Akt phosphorylation (phos), we completed DNA transfections of S6K1-GST and GST-Akt independently. After transfection, myoblasts were incubated for 2 days in normal medium. A: C2C12 myoblasts were treated with serum/antibiotic-free media for 60 min (CON), deprived of the leucine for 60 min (−leu), and restimulated for 20 min with 50 mg/ml leucine (−leu +leu). B: relative phosphorylation levels of S6K1 (T389) in each group were quantified. Values are means ± SE; n = 6 in each group. *P < 0.05, empty vector vs. REDD2-transfected groups in each experimental condition. #P < 0.05, CON vs. −leu groups in each transfection condition. †P < 0.05, −leu vs. +leu groups in each transfection condition. C: C2C12 myoblasts were plated on type I collagen-coated Bioflex membranes and grown to confluence. Myoblasts were treated with serum/antibiotic-free media for 60 min (CON) and then stimulated by 10 min of 15% biaxial stretch using a Flexcell-4000 system. D: relative phosphorylation levels of S6K1 (T389) in each group were quantified. Values are means ± SE; n = 6 in each group. *P < 0.05, empty vector vs. REDD2-transfected groups in each experimental condition. #P < 0.05, CON vs. stretch groups in each transfection condition.

Fig. 3.

Overexpression of REDD2 is not sufficient to inhibit Rheb-induced phosphorylation of S6K1 (T389). A: C2C12 myoblasts were triple transfected with S6K1-GST or GST-Akt and empty or expression vector of Rheb and/or REDD2. B: relative phosphorylation levels of S6K1 (T389) in each group were quantified. Values are means ± SE; n = 6 in each group. *P < 0.05, empty vector vs. REDD2-transfected group. #P < 0.05, empty vector vs. Rheb-transfected group.

Fig. 4.

Tuberous sclerosis complex 2 (TSC2) knockdown rescues REDD2 inhibition of mammalian target of rapamycin (mTOR) activity. For the knockdown of TSC2, predesigned small interfering (si)RNA oligonucleotides were transfected to C2C12 myoblasts. A: 3–4 days after siRNA transfection, myoblasts were treated with serum/antibiotic-free media for 60 min and deprived of leucine for 60 min. B: C2C12 myoblasts were cotransfected with S6K1-GST or GST-Akt and empty or HA-tagged REDD2. One day after plasmid DNA transfection, siRNA oligonucleotides were transfected to myoblast. Once myoblasts reached confluence, they were used for the treatment of leucine deprivation. In all experiments, knockdown efficiency of TSC2 expression was confirmed by real-time RT-PCR and Western blotting. Efficiency of TSC2 mRNA silencing was ∼90% as determined by real-time RT-PCR analysis (see Supplemental Fig. 3). For the transfection control, the Stealth RNAi Negative Control Kit (Invitrogen) was used.

Fig. 5.

REDD2 interacts with 14-3-3 protein but not with either TSC1 or TSC2. A and B: C2C12 myoblasts were cotransfected with plasmid vectors of HA-tagged REDD2 and myc-tagged TSC1 (A) or FLAG-tagged TSC2 (B). Total lysates were immunoprecipitated (IP) with each antibody. Reciprocal immunoprecipitation/immunoblotting with anti-HA and anti-myc/anti-FLAG antibodies showed that there was no direct interaction between REDD2 and TSC1 or TSC2. C and D: total cell lysates were immunoprecipitated with normal mouse IgG or pan-14-3-3 mouse monoclonal antibody. Endogenous TSC2 (C) and HA-tagged REDD2 (D) were coimmunoprecipitated with 14-3-3 antibody.

Fig. 6.

Overexpression of REDD2 induces dissociation of TSC2/14-3-3 interaction. C2C12 myoblasts were infected with increasing titer of adenovirus expressing REDD2 or Ad5-CMV-EGFP to control for virus infection. Total cell lysates were immunoprecipitated with pan-14-3-3 mouse monoclonal antibody. Increasing levels of REDD2 overexpression resulted in increased dissociation of TSC2 from its inhibitory 14-3-3 protein with no changes in levels of TSC2 or 14-3-3 protein.

Fig. 7.

Colocalization of REDD2 with TSC1 and TSC2 in the cytoplasm of C2C12 myoblasts. C2C12 myoblasts were cotransfected with plasmid vectors of HA-tagged REDD2 and myc-tagged 14-3-3β (A), myc-tagged TSC1 (B), or FLAG-tagged TSC2 (C). Intracellular colocalization of each molecule was determined by immunofluorescent analysis. Expression of REDD2 was detected with rabbit anti-HA and Texas red-conjugated goat anti-rabbit antibodies, and expression of 14-3-3β, TSC1, and TSC2 was detected with FITC-conjugated goat anti-myc or -FLAG antibodies. Visualized images were converted as monochrome with RGB splitting and then merged. Processed images were adjusted using ImageJ 1.38X software (NIH). DAPI, 4,6-diamidino-2-phenylindole.