mTOR inhibition activates overall protein degradation by the ubiquitin proteasome system as well as by autophagy Jinghui Zhao, Bo Zhai, Steven P. Gygi, and Alfred Lewis Goldberg1 Department of Cell Biology, Harvard Medical School, Boston, MA 02115 This contribution is part of the special series of Inaugural Articles by members of the National Academy of Sciences elected in 2015. Contributed by Alfred Lewis Goldberg, November 6, 2015 (sent for review September 10, 2015; reviewed by David J. Glass and David Rubinsztein)

Growth factors and nutrients enhance protein synthesis and suppress overall protein degradation by activating the protein kinase mammalian target of rapamycin (mTOR). Conversely, nutrient or serum deprivation inhibits mTOR and stimulates protein breakdown by inducing autophagy, which provides the starved cells with amino acids for protein synthesis and energy production. However, it is unclear whether proteolysis by the ubiquitin proteasome system (UPS), which catalyzes most protein degradation in mammalian cells, also increases when mTOR activity decreases. Here we show that inhibiting mTOR with rapamycin or Torin1 rapidly increases the degradation of long-lived cell proteins, but not short-lived ones, by stimulating proteolysis by proteasomes, in addition to autophagy. This enhanced proteasomal degradation required protein ubiquitination, and within 30 min after mTOR inhibition, the cellular content of K48-linked ubiquitinated proteins increased without any change in proteasome content or activity. This rapid increase in UPS-mediated proteolysis continued for many hours and resulted primarily from inhibition of mTORC1 (not mTORC2), but did not require new protein synthesis or key mTOR targets: S6Ks, 4E-BPs, or Ulks. These findings do not support the recent report that mTORC1 inhibition reduces proteolysis by suppressing proteasome expression [Zhang Y, et al. (2014) Nature 513(7518):440–443]. Several growth-related proteins were identified that were ubiquitinated and degraded more rapidly after mTOR inhibition, including HMG-CoA synthase, whose enhanced degradation probably limits cholesterol biosynthesis upon insulin deficiency. Thus, mTOR inhibition coordinately activates the UPS and autophagy, which provide essential amino acids and, together with the enhanced ubiquitination of anabolic proteins, help slow growth. mTOR

Rapamycin is a natural product that selectively inhibits mTORC1 but not mTORC2. Torin1 is a synthetic mTOR inhibitor that blocks ATP-binding to mTOR and thus inactivates both mTORC1 and mTORC2 (7). Both Torin1 and rapamycin inhibit overall protein synthesis, induce autophagosome formation, and thus mimic the effects of starvation. However, Torin1 is much more effective than rapamycin in affecting these two processes because rapamycin inhibits mTORC1 incompletely (7, 8). Because many types of cancer are associated with overactivation of mTOR, rapamycin and other novel mTOR inhibitors are useful in the treatment of certain cancers (9). Rapamycin is widely used in the clinic as an immune suppressor (10). Nevertheless, rapamycin extends lifespan in aged mice, perhaps by triggering similar changes as occur with dietary caloric restriction (11). Additionally, stimulating autophagy by rapamycin can help clear intracellular protein aggregates as they accumulate in many neurodegenerative diseases and reduce their toxicity (12). Eukaryotic cells degrade proteins by both the autophagy-lysosome system and the ubiquitin proteasome system (UPS). Macroautophagy delivers cytoplasmic proteins or organelles into autophagic vacuoles for degradation, and autophagosome formation is rapidly activated by starvation or mTOR inhibition (3). The UPS is responsible for the degradation of most cytosolic and nuclear proteins in mammalian cells, including the short-lived regulatory and misfolded proteins as well as the bulk of cell constituents, which are long-lived components (13, 14). Degradation by the UPS is highly selective, involving the Significance Lack of nutrients or growth factors rapidly inhibits mTOR kinase and stimulates overall protein breakdown by autophagy, which provides amino acids for new protein synthesis and energy production. Here, we demonstrate that mTOR inhibition also rapidly increases overall protein degradation by the ubiquitin proteasome system by enhancing the ubiquitination of many cell proteins. These findings demonstrate a new general mechanism regulating overall protein breakdown in mammalian cells. This rapid activation of proteolysis not only slows growth and provides essential amino acids upon starvation, but also stimulates ubiquitin-dependent degradation of certain growth-related proteins, including HMGCoA synthase, which is critical for cholesterol biosynthesis. Accelerated degradation of such proteins represents a new mechanism for rapid suppression of anabolic processes when mTORC1 activity decreases.

| proteasome | ubiquitination | autophagy

T

he balance between overall rates of protein synthesis and degradation determines whether a cell grows or atrophies. It has long been proposed that overall rates of protein synthesis and degradation can be coordinately regulated (1). For example, when hormones (e.g., insulin and insulin-like growth factor-1) and nutrients are plentiful, rates of protein synthesis are high and protein degradation is suppressed, whereas in starving cells, synthesis falls and overall degradation rises. One critical factor coordinating overall synthesis and degradation is the Ser/Thr protein kinase mammalian target of rapamycin (mTOR), which promotes protein translation (2) while suppressing autophagy (lysosomal proteolysis) (3). mTOR’s pleiotropic functions are catalyzed by two distinct kinase complexes: mTORC1, the key component of which is Raptor, and mTORC2, which instead contains Rictor (4, 5). Growth factors act through class I PI3K and Akt kinases to inhibit the tumor suppressor TSC2, and thereby activate mTORC1 (6). Adequate supply of amino acids, especially leucine, can also activate mTORC1, but through a distinct mechanism involving Rag GTPase and lysosomal recruitment of mTORC1 (4). Unlike mTORC1, mTORC2 is insensitive to amino acid supply but is activated by growth factors via mechanisms that remain unclear.

15790–15797 | PNAS | December 29, 2015 | vol. 112 | no. 52

Author contributions: J.Z. and A.L.G. designed research; J.Z., B.Z., and S.P.G. performed research; J.Z. and A.L.G. analyzed data; and J.Z. and A.L.G. wrote the paper. Reviewers: D.J.G., Novartis Institutes for Biomedical Research; and D.R., University of Cambridge. The authors declare no conflict of interest. 1

To whom correspondence should be addressed. Email: Alfred_Goldberg@hms. harvard.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1521919112/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1521919112

Results Inhibiting mTOR Rapidly Enhances the Degradation of Long-Lived but Not Short-Lived Cell Proteins. We first tested if mTOR inhibitors affect

the degradation of radiolabeled short-lived and more stable proteins in HEK293 cells by using pulse-chase approaches described previously (14). Treatment with either Torin1 or rapamycin enhanced the degradation of long-lived cell proteins (labeled for 20 h) reproducibly by 1 h, although Torin1 consistently caused a larger increase (Fig. 1B). In contrast, neither agent increased the breakdown of short-lived proteins, which were degraded within 1 h after a brief labeling (20 min) (Fig. 1A). Therefore, subsequent studies focused on the degradation of long-lived proteins, which comprise the great majority of cell proteins. Torin1 inactivates both mTORC1 and mTORC2 (7), as was confirmed by the elimination of phosphorylated S6K (Thr389) and Akt (Ser473) (Figs. 2B, 3F, and 4C). However, rapamycin, which inhibits only mTORC1, abolished S6K phosphorylation, while causing hyperphosphorylation of Akt (Fig. 2B and SI Appendix, Fig. S6). Because rapamycin inhibits mTORC1 only partially (7), we mainly used Torin1 to clarify mTOR’s role, but also studied rapamycin because of its wide use in research and the clinic. Inhibiting mTOR Stimulates both Autophagy and Proteasome-Mediated Proteolysis. To determine whether mTOR inhibitors enhance protein

breakdown by the UPS, we measured their effects on the degradation mediated by proteasomes or lysosomes by analyzing the fraction of overall proteolysis sensitive to the proteasome inhibitor, bortezomib (BTZ), or to the inhibitor of lysosomal acidification, concanamycin A (CCA), as described previously (14). As expected, Torin1 and to a lesser extent rapamycin increased lysosomal (CCA-sensitive) proteolysis in both HEK293 cells and mouse embryonic fibroblasts (MEFs) (Fig. 1D). The larger stimulation of lysosomal proteolysis by

Fig. 1. mTOR inhibitors enhance the degradation of SHORT-LIVED LONG-LIVED 8 long-lived, but not short-lived, cell proteins by actiTorin1 20 vating proteasomal proteolysis, as well as autophagy. (A) Rapamycin and Torin1 did not affect the degraRapamycin 6 15 dation of short-lived cell proteins in HEK293 cells. Control Newly synthesized proteins were radiolabeled by in4 10 cubating cells in complete medium supplemented with 3H-Phe for 20 min. After two quick washes of the 2 5 cells with medium containing 2 mM nonradioactive Phe and cycloheximide to prevent reuptake and rein0 0 corporation, control (♦, vehicle), rapamycin (□, 300 nM), 0 0.5 1 1.5 3 6 1 5 0 2 4 or Torin1 (▲, 100 nM) were added at time 0, and the Hours Hours release of 3H-Phe from cell proteins into the media was measured especially during the first hour when short-lived proteins are selectively degraded. Proteasomal degradation Lysosomal degradation (B) Rapamycin and Torin1 increased the degradation * * Control 1.0 * of long-lived proteins in HEK293 cells. To label long1.2 * Rap * * lived proteins selectively, cells were incubated with 3H0.8 * * Torin1 1.0 0.8 * Phe for 20 h. After switching to chase medium for 2 h * to allow the degradation of short-lived components, 0.8 0.6 0.6 fresh chase medium containing mTOR inhibitors was 0.6 * added, and proteolysis assayed. The fraction of ra0.4 0.4 0.4 dioactive proteins degraded to TCA-soluble material * 0.2 0.2 at different times was expressed as a percentage of 0.2 total radioactivity incorporated into cell proteins at t 0 0 0 = 0. Significant increases upon mTOR inhibition were HEK293 MEF MEF MEF HEK293 MEF MEF observed at 90 min and thereafter. (C) mTOR in(Atg5+/+) (Atg5−/−) (Atg5−/−) (Atg5+/+) (Atg5−/−) hibition with rapamycin or Torin1 increased degradation by proteasomes, which was measured as CCA-resistant proteolysis in HEK293 cells and MEFs (Left), or by BTZ-sensitive proteolysis in Atg5−/− MEFs (Right). (D) mTOR inhibition increased lysosomal (CCA-sensitive) proteolysis in HEK293 cells and wild type MEFs, but not in Atg5−/− MEFs. Error bars (SEM) were provided in C and D and for every time point in A and B (but were hidden by labels in B). *P < 0.05.

B

% Protein degraded

% Protein degraded

A

CCA-sensitive proteolysis (%/h)

D

BTZ-sensitive proteolysis (%/h)

CCA-resistant proteolysis (%/h)

C

Zhao et al.

INAUGURAL ARTICLE

conditions (18). In this study, we have used well-validated pulsechase approaches (14) to measure proteolysis by proteasomal and lysosomal systems after mTOR inhibition and have also monitored overall Ub conjugation and proteasome content.

PNAS | December 29, 2015 | vol. 112 | no. 52 | 15791

BIOCHEMISTRY

attachment of a ubiquitin (Ub) chain to the substrate through the sequential actions of a Ub-activating enzyme (E1), Ub-conjugating enzymes (E2s), and one of the cell’s many Ub ligases (E3s) (15). Proteins conjugated to Ub chains are rapidly degraded by the 26S proteasome, whereas the Ub molecules are recycled by proteasome-associated deubiquitinating enzymes (DUBs) (16). Although the degradation of specific short-lived proteins has been studied extensively, our understanding of the global mechanisms controlling proteasomal degradation of the bulk of cell constituents is very limited. The best-characterized function of mTOR is to enhance protein translation through mTORC1-mediated phosphorylation of 4E-BPs and S6Ks (2). The simultaneous suppression of protein degradation is generally attributed to the ability of mTORC1 to inhibit autophagy, which seems to occur by the phosphorylation and inactivation of Atg1/Ulks (3). The goal of this study was to test whether the increased proteolysis upon mTOR inhibition may also occur through activation of the UPS. We demonstrate herein that mTOR inhibition not only enhances lysosomal proteolysis, but also rapidly stimulates the ubiquitination and proteasomal degradation of many proteins. We also investigated the mechanisms by which the UPS is activated, the role of mTORC1 or mTORC2, and the nature of the proteins whose stability is affected by mTOR. Together with the enhancement of autophagy, this activation of proteolysis by the UPS upon mTOR inhibition appears to represent an important adaptation to starvation that helps slow growth and provide essential amino acids. A very different role of mTORC1 in regulating protein degradation was recently proposed by Zhang et al. (17), who surprisingly reported that rapamycin did not rapidly increase proteolysis, but after 16 h actually reduced overall proteolysis by decreasing proteasome expression (17). Such a suppression is inconsistent with the well-established increase in proteolysis and autophagy in starving cells, and various observations about the regulation of the UPS (see below). Zhang et al.’s conclusions probably result from their potentially misleading pulsechase methods to measure protein degradation and cell culture

C

1.0

B

Control Torin1

*

Rap: Torin1:

+ −

+ −

+ −

− −

− −

Activation of Proteasomal Degradation upon mTOR Inhibition Does Not Require Protein Synthesis. Although mTOR inhibition alters

rates of translation (8) and transcription (20, 21), it rapidly stimulates autophagy independently of protein synthesis (3). To learn whether the enhancement of proteasomal degradation by mTOR inhibitors requires protein synthesis, we pretreated HEK293 cells with cycloheximide for 1 h before Torin1 addition. Although cycloheximide by itself reduced proteasomal proteolysis [probably by enhancing mTOR activity (22)] after cycloheximide treatment, Torin1 caused a similar or greater increase in proteasomal proteolysis (Fig. 2A). Thus, this rapid stimulation by mTOR inhibitors does not require new protein or mRNA synthesis. mTOR Inhibition Rapidly Stimulates K48-Linked Ub Conjugation to Proteins in Cells and in Vivo, Which Leads to Increased Proteolysis.

For most proteins, ubiquitination is the rate-limiting step for degradation by the UPS. We therefore tested if inhibiting mTOR enhances ubiquitination generally. Because newly synthesized proteins account for a large fraction of the ubiquitinated proteins in cells and mTORC1 influences their rates of synthesis, we pretreated MEFs with cycloheximide for 1 h to focus on the ubiquitination of long-lived proteins. Although cycloheximide decreased the total content of ubiquitinated proteins (Fig. 2C and SI Appendix, Fig. S11), addition of rapamycin or Torin1 for 1 h resulted in a highly reproducible 30% increase in the total content of ubiquitinated proteins and a concomitant 30% reduction in free Ub levels (Fig. 2B). This increase in Ub conjugates was maximal after only a 30-min treatment and persisted

− −

− +

− +

− +

*

0.8

188

0.6

Ub(P4D1):

98

0.4

62

0.2 0

Ub(P4D1):

8

P-Akt(473): Vehicle

CHX

P-S6K(389): β-actin:

CHX: Rap: Torin1:

− + −

− − −

− − +

+ + −

+ + −

+ − −

+ − −

+ − +

+ − + 188

Ub (K48):

98

188

Ub (K63):

98

120 100

**

80 60 40 20 0

62

Control Rap Torin1

**

140

Relative abundance

A

Proteasomal proteolysis (%/h)

Torin1 is consistent with morphological evidence that Torin1 is more effective than rapamycin in inducing autophagosome formation (7). These increases in lysosomal degradation depend on autophagy, because Torin1 failed to stimulate this process in MEFs unable to form autophagosomes because of deficiency of Atg5 (Fig. 1D). In addition, Torin1 and rapamycin increased proteasomal degradation, measured by either of two approaches: the CCAresistant proteolysis in HEK293 cells and MEFs (Fig. 1C, Left) or the BTZ-sensitive proteolysis in Atg5-null MEFs (Fig. 1C, Right). In autophagy-competent cells, the CCA-resistant component was used to measure proteasomal degradation (Fig. 1C, Left) because we found that although BTZ does not directly inhibit lysosomal function (19), it partially reduced the activation of autophagy by Torin1 (SI Appendix, Fig. S1) (probably by blocking the degradation of an inhibitor of autophagy). Torin1 treatment similarly increased proteasomal proteolysis in both undifferentiated and differentiated C2C12 cells (SI Appendix, Fig. S2). Therefore, mTOR inhibition in many cell types stimulates overall proteolysis independently of autophagy. In several cell types analyzed, proteasomes accounted for twothirds or more of the degradation of long-lived proteins, and lysosomes for nearly all of the remainder (SI Appendix, Table S1). Upon mTOR inhibition, although the relative increases in lysosomal proteolysis are much larger than that by the UPS (especially with Torin1), proteasomes still accounted for a greater or equal fraction of the total proteolysis than lysosomes (Fig. 1 C and D).

Ub Free Ub SUMO2/3 Conjugates Conjugates

β-actin:

Control Rap

Mouse liver

− − − + + +

140

188

Ub:

98

(FL-76) 62

β-actin:

Levels of Ub-conjugates

Rap:

120

*

100 80 60 40

E Normalized peptidase activity of 26S

D

80 60 40 20 0

20 0

15792 | www.pnas.org/cgi/doi/10.1073/pnas.1521919112

Control Torin1

100

α-subunits:

Lysate

Purified

Fig. 2. mTOR inhibition rapidly increases the levels of K48-linked Ub conjugates in cells and mouse liver without affecting proteasomal activity. (A) The increase in proteasomal degradation by Torin1 treatment in HEK293 cells did not require protein synthesis. Proteasomal proteolysis was determined in the presence or absence of 20 μg/mL cycloheximide. (B) mTOR inhibition increased the cellular content of Ub conjugates but not SUMO2/3 conjugates in MEFs, which were pretreated with cycloheximide for 1 h and then with rapamycin or Torin1 for 1 h. Ub conjugates, free Ub, and SUMO2/3 conjugates were determined by Western blot, and their intensities quantitated and normalized to that of β-actin. Phospho-S6K and -Akt were measured to confirm the efficacy of rapamycin and Torin1 treatments. (C) mTOR inhibition raised K48linked, but not K63-linked, Ub conjugates in MEFs. MEFs were pretreated with cycloheximide for 1 h and then treated with rapamycin or Torin1 for 1 h. K48 or K63-linked conjugates were determined by Western blot with linkage-specific antibodies, and their intensities quantitated and plotted in SI Appendix, Fig. S11. (D) Rapamycin raised the level of Ub conjugates in mouse liver. CD1 mice (body weights ∼35 g) were injected intraperitoneally with 2.5 mg/kg rapamycin or PBS. 2 h later, mice were killed, livers homogenized, and 100 μg total proteins were analyzed by Western blot. (E) Torin1 treatment did not increase peptidase activity of 26S proteasomes. HEK293 cells were first treated with Torin1 or vehicle for 1 h before lysis, and activity of 26S proteasomes in lysates was measured by using the fluorogenic substrate Suc-LLVY-amc. Second, we treated a mouse myoblast cell line that expresses a FLAG-tagged PSMB2 with Torin1 or vehicle for 1 h before lysis. 26S proteasomes were then isolated with a FLAG-antibody resin, and peptidase activity of the purified 26S proteasomes measured. Western blots of α-subunits of the proteasome are included to show equal loading. *P < 0.05.

Zhao et al.

0.8

β-actin:

0

0.10

*

−

*

*

0.08 0.06

0.06 0.04

0.04

0.02

0.02

0

0

Akti-1/2

Vehicle

+

D

Torin1

SUPT6H: P-S6K(389): P-Pras40: β-actin:

G *

Control Torin1

*

Proteasomal 0.4 0.3

0.6

HMGCS1:

0.4

SUPT6H:

0.2

α-taxilin:

*

*

*

0.8 0.6

0.2

0.4

0.1

0.2

0

80

*

60 40 20 0

0

I

100

* HMGCS1 SUPT6H

Rictor +/+ Rictor −/−

0.8

Control Akti-1/2 Torin1

1.0

H Rictor: P-Akt(473): P-S6K(389):

E

Lysosomal

0.5

Proteasomal proteolysis (%/h)

+

Proteasomal proteolysis (%/h)

*

−

HMGCS1:

Proteasomal proteolysis (%/h)

INAUGURAL ARTICLE 62

0.08

*

n.s. n.s.

1.2 1.0 0.8 0.6 0.4 0.2

0 Torin1: + Insulin: −

− −

Medium: Complete P-S6K(389):: P-Akt(473):: Akt::

β-actin: Rictor+/+ Rictor−/−

for at least 3 h (SI Appendix, Fig. S3B). Similar increases in Ub conjugate levels were seen in HEK293 cells using three different Ub antibodies (FL-76, FK2, P4D1) (SI Appendix, Fig. S3A) and no change was observed in the levels of SUMO2/3 conjugates (Fig. 2B and SI Appendix, Fig. S3C). Ub chains are formed through isopeptide linkages between the C-terminal glycine of Ub and one of the lysines on the preceding Ub. Although K48-linked Ub chains target proteins to 26S proteasomes, K63-linked chains do not (23). Using antibodies specific for these two linkage types, we found that rapamycin and Torin1 increased the cellular content of K48-linked Ub chains, but not of K63-chains (Fig. 2C and SI Appendix, Fig. S11). To learn whether mTOR inhibition also increases the level of Ub conjugates in vivo, we injected rapamycin into mice intraperitoneally and removed the livers 2 h later. An ∼30% similar increase in Ub conjugate levels was observed in liver extracts (Fig. 2D). This rapid increase in K48-linked Ub conjugates can account for the enhanced degradation by proteasomes after mTOR inhibition and must reflect increased ubiquitination of many cell proteins. Accordingly, treating MEFs with a specific inhibitor of the major Ubactivating enzyme Ube1 (24) depleted the MEFs of nearly all Ub conjugates in 1 h (Fig. 3B) and reduced proteasomal proteolysis by at least 60% (Fig. 3A). Importantly, inhibition of Ube1 completely eliminated the enhancement of proteasomal degradation by Torin1 Zhao et al.

98

(FK2)

0.2

F6h + CHX:

0

188

0.4

0.10

1.0

+

Ub:

0.12

1.2

−

Ube1 inhi:

Control Torin1

*

Rate of proteolysis (%/h)

Increase in proteasomal proteolysis by Torin1 (%/h)

*

0.6

Ube1 inhi:

C

B

Lysosomal

*

Relative protein levels

Rate of proteolysis (%/h)

Proteasomal

− −

− + EBSS

+ +

Fig. 3. The increased proteasomal degradation by Torin1 requires ubiquitination, but not Akts, S6Ks, and Ulks, and results primarily from mTORC1 inhibition. (A) The increase in proteasomal degradation induced by mTOR inhibition requires ubiquitination. Proteasomal and lysosomal proteolysis were determined in the presence or absence of 10 μM Ube1 inhibitor in MEFs. (B) Blocking ubiquitination markedly reduced the cellular content of Ub conjugates. MEFs were treated with the Ube1 inhibitor or vehicle for 1 h before lysis, and the levels of Ub conjugates was measured by Western blot with anti-Ub (FK2). (C) The increase in proteasomal proteolysis by Torin1 was independent of S6Ks or 4E-BPs. Proteasomal proteolysis was determined in wild-type MEFs and MEFs lacking 4E-BP1 and 2 or S6K1 and 2. (D) Overexpression of Ulk1 or Ulk2 enhanced lysosomal but not proteasomal degradation. Proteolysis was determined in HEK293 cells after 20 h of transfection with vectors expressing Ulk1, Ulk2, or an inactive mutant of Ulk1 (Ulk1-K46N). (E) Unlike Torin1, the Akt inhibitor, Akti-1/2, (2 μM) did not increase proteasomal proteolysis in MEFs. (F) Akt inhibition did not induce the degradation of HMGCS1 and SUPT6H. The contents of these proteins were measured by Western blot after treatments for 6 h in the presence of cycloheximide. PhosphoPras40 was measured to confirm the efficacy of Akti-1/2. (G) Torin1 increased proteasomal degradation in mTORC2-deficient (Rictor-null) MEFs. (H) Rictor-null MEFs showed no mTORC2 activity, reduced mTORC1 activity, and similar levels of HMGCS1, SUPT6H, and α-taxilin proteins compared with wild-type MEFs. (I) Selective activation or inhibition of mTORC2 did not affect proteasomal degradation in HEK293 cells. Proteasomal proteolysis was measured as described in Methods. For EBSS treatment, cells were washed once with EBSS and incubated with EBSS plus or minus insulin for 30 min, and then Torin1 was added for 1 h before measuring proteolysis. *P < 0.05; n.s., not significant.

(Fig. 3A). Although blocking ubiquitination also reduced slightly the increase in lysosomal proteolysis by Torin1, about 70% of the Torin1induced activation of lysosomal degradation was independent of ubiquitination (Fig. 3A). Thus, the increased protein ubiquitination after mTOR inhibition is required for the accelerated proteasomal degradation of long-lived proteins. We also tested for a possible additional effect of Torin1 in stimulating proteasomal activity, but could detect no change in 26S proteasomal peptidase activity in HEK293 extracts or after purification of proteasomes from the treated cells (Fig. 2E). In addition, we did not detect any effect of mTOR inhibitors on overall deubiquitinating activity in HEK293 extracts, measured using the model substrate, Ub-amc (SI Appendix, Fig. S4A), nor did we observe any change in the activities of many individual DUBs measured by their modification with the active site probe, HA-Ub-vinyl sulfone (SI Appendix, Fig. S4B). mTOR Inhibition Increases Ubiquitination and Degradation of GrowthRelated Proteins. To identify some of the long-lived proteins whose

ubiquitination is stimulated by mTOR inhibition, we pretreated HEK293 cells with cycloheximide and then with Torin1 or vehicle for 1 h. We then isolated ubiquitinated proteins from the lysates using immobilized tandem-repeated Ub-binding entities (25) and analyzed the conjugates by mass spectrometry (Fig. 4A). Many proteins were identified but showed little or no change in their PNAS | December 29, 2015 | vol. 112 | no. 52 | 15793

BIOCHEMISTRY

A

B

Lysis of HEK293 cells treated with vehicle or Torin1 for 1 hour

Spectral counts by MS Protein names

Enrichment of ubiquitinated proteins by immobilized GST-TUBE

2

9

9

SUPT6H

3

7

2

5

α-taxilin

8

13

3

11

Myst2

0

3

1

4

Identification of ubiquitinated proteins by mass spectrometry

Torin1: BTZ: CCA: MLN4924:

− − − −

− − − −

− − − −

+ − − −

Exp-2

HMGCS1 Release of ubiquitinated proteins from GST by PreScission protease cleavage

C

Exp-1

Control Torin1 Control Torin1

+ − − −

+ + − −

+ − − −

+ − + −

+ − − +

− + − −

− − + −

HMGCS1: SUPT6H: α-taxilin: Myst2: P-S6K(389):

18

CHX 6h CHX+Torin 6h

120

Relative protein levels

A

100

*

80

#

#

60 40 # 20 0

P-Akt(473): β-actin:

D

BTZ

3

6

Torin1 T+BTZ

EBSS

E+BTZ

3

3

3

6

3

6

6

F Insulin/IGF-1

6 (h)

HMGCS1: 1.0 1.0 1.0 0.7 0.4 0.9 0.9 0.6 0.2 1.0 1.0

β-actin:

Amino Acids

PI3K

E

Proteasomal proteolysis (%/h)

AKT

0.8

*

*

Control Torin1

mTOR

0.6 0.4 0.2 0

Vehicle MLN4924

Protein Autophagy Ubiquitination & synthesis Proteasomal proteolysis

ubiquitination following mTOR inhibition, as determined by counts of peptides recovered (SI Appendix, Table S2). However, we identified 55 proteins for which at least two peptides were detected (see examples in Fig. 4B; SI Appendix, Table S3) that showed at least a 50% increase in ubiquitination after Torin1 treatment in two independent experiments and we tested whether mTOR inhibition also enhances their degradation. Antibodies were available for 12 of the proteins that showed increased ubiquitination, and Torin1 treatment in the presence of cycloheximide reproducibly decreased the levels of four proteins, including HMGCS1 (cytoplasmic HMGCoA synthase), SUPT6H, α-taxilin, and Myst2 (Fig. 4C). Their accelerated degradation was mediated by proteasomes because these decreases were blocked by BTZ, but not by CCA (Fig. 4C). For the other eight proteins tested, mTOR inhibition did not significantly alter their degradation within 6 h (SI Appendix, Fig. S5). Among these four proteins, the degradation of HMGCS1 was stimulated to the largest extent by Torin1, which decreased its half-life from >10 h to about 2.5 h (SI Appendix, Fig. S6). In the mevalonate pathway for synthesis of cholesterol and isoprenoids, HMGCS1 precedes HMG-CoA reductase (26). Inhibition of mTORC1 reduces the transcription of HMGCS1 and other enzymes in this pathway through SREBP2 (sterol regulatory elementbinding protein 2) (20, 27, 28), and we found that Torin1 also rapidly decreased the synthesis of HMGCS1 by 25% (SI Appendix, Fig. S7). Therefore, mTOR inhibition reduces levels of HMGCS1 15794 | www.pnas.org/cgi/doi/10.1073/pnas.1521919112

Fig. 4. Identification of proteins that are ubiquitinated and degraded more rapidly after Torin1 treatment. (A) Protocol for identifying proteins with increased ubiquitination after Torin1 treatment of HEK293 cells. (B) Examples of the 55 proteins with increased ubiquitination after Torin1 treatment, as shown by increased number of peptides identified by mass spectrometry of ubiquitinated proteins. (C) Torin1 treatment increased the proteasomal degradation of 4 of the 12 proteins for which antibodies were obtained. Western blot of cell extracts after treatments of MEFs for 6 h in the presence of cycloheximide, and their levels were quantified. (D) The marked reduction in HMGCS1 protein levels after treatment with Torin1 or EBSS buffer for 3–6 h was prevented by BTZ treatment. (E) Torin1-caused increase in proteasomal proteolysis did not require a cullin-based Ub ligase. MLN4924 (2 μM) was used to inactivate all cullin-based Ub ligases. (F) Summary of the regulation of proteasomal and autophagic proteolysis by mTOR, downstream of nutrient and growth factor signaling. *P < 0.05; #P < 0.01.

by three mechanisms: decreased transcription and translation, which affect enzyme levels only slowly (because of its long half-life), and by accelerated degradation, which allows enzyme levels to change much faster. Accordingly, we observed a marked reduction in HMGCS1 levels upon Torin1 treatment or deprivation of growth factors and amino acids (EBSS buffer) for 3–6 h that could only occur through enhanced degradation, not by reduced synthesis. In fact, BTZ completely prevented this fall in HMGCS1 content by Torin1 or EBSS buffer (Fig. 4D). This finding of accelerated degradation of HMGCS1 after mTOR inhibition represents a new mechanism for rapid inhibition of the biosynthesis of cholesterol and isoprenoids upon nutrient or insulin deficiency. The other three proteins also seem likely to have functions important for cell growth: transcription elongation and mRNA processing for SUPT6H, vesicular trafficking for α-taxilin, and histone acetylation for Myst2. The contents of α-taxilin and Myst2 are elevated in several tumors (29–32) and some proliferating normal cells (33, 34). However, we could only examine the stabilities of 12 proteins, and many ubiquitinated proteins must have been missed by this approach because of their low abundance, rapid degradation, or incomplete isolation. Most likely, these four represent only a small fraction of the proteins whose ubiquitination and degradation are suppressed by mTOR. Stabilizing key anabolic proteins along with the reduced ubiquitination of cell Zhao et al.

The Enhancement in Proteasomal Proteolysis Does Not Involve mTOR Targets 4E-BPs, S6Ks, Ulks, or Akt. Because the increases in

proteasomal proteolysis and autophagy upon mTOR inhibition occur simultaneously with the decrease in protein synthesis, we investigated whether the mTOR targets that activate translation or inhibit autophagy might also regulate proteolysis by the UPS. S6Ks and 4E-BPs mediate the stimulation of translation by mTORC1 (2). However, neither appears to be important for effects on proteasomal proteolysis, because Torin1 treatment of MEFs lacking S6K1 and S6K2 or lacking 4E-BP1 and 4E-BP2 increased proteasomal proteolysis as in wild-type MEFs (Fig. 3C). Ulk1 and Ulk2 are critical for the suppression of autophagy by mTORC1 (3). Although overexpression of either Ulk1 or Ulk2, but not the catalytically inactive mutant Ulk1-K46N, stimulated proteolysis by lysosomes, these treatments did not affect proteasomal degradation (Fig. 3D). Because Akts are substrates of mTORC2 and activate many growthrelated processes, we tested if Akts are also important in suppressing proteasomal proteolysis. However, inhibition of Akts with Akti-1/2, as confirmed by the loss of phosphorylated Pras40 (Thr246) (Fig. 3F), did not increase proteasomal proteolysis generally (Fig. 3E), nor the degradation of HMGCS1 and SUPT6H (Fig. 3F). Although the simultaneous stimulation by mTOR of translation and the inhibition of autophagy and proteasomal proteolysis are coordinated responses that together promote protein accumulation, they are signaled by distinct mechanisms. The Increase in Proteasomal Degradation Results Mainly from Inhibition of mTORC1. To determine whether the enhancement

of proteasomal proteolysis by Torin1 is caused by an inhibition of mTORC1 or mTORC2, we performed two types of experiments. First, in Rictor-null MEFs that lack mTORC2 activity (Fig. 3H), Torin1 treatment still increased overall proteasomal degradation (Fig. 3G) and enhanced HMGCS1 degradation as in wild-type MEFs (SI Appendix, Fig. S8). Thus, mTORC1 alone stabilizes many cell proteins. Second, when HEK293 cells were deprived of both serum and amino acids, which inactivated completely mTORC1 activity and reduced significantly mTORC2 activity, proteasomal proteolysis increased (Fig. 3I). However, addition of insulin, which activated strongly mTORC2 but not mTORC1, did not reduce the enhancement of proteasomal proteolysis, and inhibiting this hyperactivation of mTORC2 by Torin1 did not enhance it either (Fig. 3I). Taken together, the ability of Torin1 to activate proteasomal proteolysis in the absence of mTORC2 (Fig. 3G) and the lack of effect of mTORC2 activation or inhibition alone (Fig. 3I) indicate that in HEK293 cells and MEFs, mTORC1 plays the major role in suppressing overall degradation by the UPS. Discussion This New General Mechanism Controlling Protein Turnover and the Conflicting Report. The present studies demonstrate that in nutrient-

rich environments and during growth, mTOR (primarily mTORC1) decreases overall protein degradation by the UPS and stabilizes many normally long-lived proteins by suppressing their ubiquitination. Under these conditions, mTOR does not seem to affect the breakdown of short-lived proteins (i.e., misfolded proteins or highly regulated proteins, whose ubiquitination has been studied most extensively). When mTOR levels fall, upon lack of nutrients or insulin, this mechanism can rapidly enhance (within 30–60 min) the supply of amino acids without requiring alterations in transcription or translation. As we were preparing this report, opposite findings were reported by Zhang et al. (17), who concluded that rapamycin treatment for 16 h or longer reduces overall proteolysis by decreasing proteasome expression through decreases in the expression of the transcription factor, Nrf1 (17). Surprisingly, no change in rates of proteolysis was Zhao et al.

The Potential Mechanisms by Which mTORC1 Suppresses Proteasomal Degradation. Our various experiments indicate that the activation by

mTOR inhibition of protein breakdown by the UPS results primarily from inhibition of mTORC1 and is driven by increased ubiquitination without any demonstrable activation of the proteasomes. In several prior reports, mTOR-mediated phosphorylation of proteins was shown to influence their ubiquitination and stability. For example, Grb10, a negative regulator of growth factor signaling, was reported to be stabilized by mTORC1 (39, 40). Among the four proteins we identified that show accelerated degradation upon mTOR inhibition, SUPT6H was reported to be a potential substrate of mTOR. Among the 55 proteins we identified that show increased ubiquitination after mTOR inhibition, 9 were reported to be tentative mTOR substrates (39, 40). Therefore, mTORC1-mediated phosphorylation of certain proteins may lead to their stabilization by inhibiting their ubiquitination, and this mechanism may contribute to the general increases in Ub-dependent proteolysis upon mTOR inhibition. Another possibility is that mTOR directly inhibits certain ubiquitinating enzymes or stimulates DUBs, and thus influences the degradation of some proteins. However, we have not succeeded in finding such enzyme PNAS | December 29, 2015 | vol. 112 | no. 52 | 15795

INAUGURAL ARTICLE

observed until 16 h of rapamycin treatment, in sharp contrast to the rapid increase in proteasomal degradation described here and the rapid activation of autophagy observed by many prior investigators (3). Their proposed mechanisms are also inconsistent with a number of prior findings about the UPS. Proteasomes are long-lived cell constituents (subunit half-lives of 40–200 h) (35). Thus, simply decreasing proteasome gene transcription should not reduce their content within a short period. In fact, after rapamycin treatment for 24 h, we did not detect any significant decrease in the levels of several proteasome subunits (SI Appendix, Fig. S9) or in 26S activity (peptidase and Ub conjugate degradation) (Fig. 2E). Furthermore, proteasomes appear to be present in excess in cells (36), and their amount is usually not rate-limiting for proteolysis. In fact, we demonstrated recently that a modest (∼30%) reduction of 26S proteasome content by knocking down PSMD2 has no effect on overall protein degradation by proteasomes (37). Most importantly, using the same cells they studied, we did not find any decrease in proteolysis, but only that mTOR inhibition activates overall proteolysis rapidly and for at least 28 h (18). As discussed elsewhere (18), both the pulse-chase analysis and experimental design of Zhang et al. (17) appear potentially misleading. In our experiments, we always maintained cells in complete medium and compared the effects of mTOR inhibition on degradation of the exact same pool of radiolabeled proteins. However, their experiments compared the breakdown of proteins labeled in TSC2-null, TSC2-expressing, or rapamycin-treated MEFs, where patterns of transcription and translation differ as a result of differences in mTORC1 activity (17), and thus, they have compared the degradation of distinct sets of proteins and not just study mTORC1’s effects on the proteolytic machinery. Additionally, Zhang et al.’s experiments involved depriving TSC2-null MEFs of serum for extended periods (up to ∼66 h) (17) to study mTORC1’s actions separately from mTORC2’s. However, this approach is problematic for studies of proteolysis because serum deprivation by itself stimulates proteolysis in all cells, and based on the present findings should do so more in wild-type cells than in TSC2-null MEFs. In fact, when we measured the breakdown of the same set of proteins in TSC2-null MEFs cells after serum deprivation for 17 h, the subsequent inhibition of mTORC1 by rapamycin or Torin1 still increased overall proteolysis (SI Appendix, Fig. S10), as we consistently found in complete medium. Furthermore, long-term mTOR inhibition or serum deprivation may trigger secondary responses. In fact, although serum deprivation acutely activates autophagy, longterm serum deprivation or sustained inactivation of insulin-like growth factor 1 signaling reduces autophagy (38). Therefore, we focused in this study on the effects of mTOR inhibition for minutes or several hours.

BIOCHEMISTRY

proteins generally seem to represent additional new mechanisms by which mTOR promotes growth.

candidates. Possibly, multiple mechanisms downstream of mTORC1 (e.g., phosphorylation of substrates, Ub ligases, or DUBs) together account for the increases in ubiquitination and proteasomal proteolysis upon mTOR inhibition. It will clearly be important to learn which Ub ligases are responsible for the increased ubiquitination of the mTOR-regulated proteins, such as HMGCS1. A large fraction of the Ub ligases in the human genome belong to cullin-based protein complexes, many of which modify phosphorylated proteins selectively. However, these cullin-based Ub ligases appear not to be involved, because exposing cells to the selective neddylation inhibitor MLN4924 (41), which inhibits their activities, did not inhibits the degradation of these four proteins (Fig. 4C), nor the increase in overall proteasomal proteolysis by Torin1 (Fig. 4E). Thus, the responsible Ub ligases belong to the monomeric RING or HECT family. Although cells contain a family of proteins that bind ubiquitinated proteins and can target them to autophagic vacuoles for degradation (42), it seems most likely that a great majority of proteins showing increased ubiquitination upon mTOR inhibition are degraded by the 26S proteasome. As shown in Fig. 3A, the Torin1-induced increase in lysosomal proteolysis was largely unaffected by blocking Ub conjugation, whereas the increase in proteasomal proteolysis required ubiquitination. In addition, the four proteins identified as undergoing accelerated ubiquitination upon Torin1 treatment were all degraded by proteasomes and not lysosomes (Fig. 4C). Thus, the stimulation of autophagy and proteasomal degradation differs in their dependence upon ubiquitination, and the increase in protein ubiquitination probably drives the enhanced proteasomal proteolysis. Although the degradation of many individual proteins by the UPS has been extensively studied, global regulation of proteolysis by this pathway has received little attention, even though a general stimulation of degradation by the UPS is critical in catabolic states, such as muscle atrophy (43). Recently, a very different type of global regulation of the UPS has been described in which proteasome function is activated by cAMP and PKAdependent phosphorylation, where proteolysis rises without any increase in total ubiquitination (44, 45). It is noteworthy that these two very different biochemical mechanisms also have distinct physiological roles and substrates. The PKA-mediated phosphorylation of proteasomes stimulates selectively the degradation of short-lived regulatory or misfolded proteins (e.g., aggregation-prone proteins causing neurodegenerative diseases) (44, 45), whereas mTOR affects the breakdown by the UPS of long-lived components, which comprise the bulk of cell proteins and thus must be important in growth regulation and adaptation to starvation. Coordinate Regulation of the UPS and Autophagy in Growing and Starving Cells. Traditionally, the UPS and the lysosomal pathway

have been assumed to function independently, and autophagy activation has been viewed as a compensatory mechanism when proteasome function is inhibited (46). However, our prior studies demonstrated that in fasting, after denervation and in other catabolic states, the loss of muscle protein (atrophy) involves coordinate activation of these two degradative systems through the activation of FoxO transcription factors (14, 47). Unlike mTOR inhibition, which enhances proteolysis via the UPS and autophagy within minutes, the FoxO-mediated stimulation requires hours for transcription and translation, but can lead to prolonged activation of these two degradative systems (14, 47). Most likely, upon starvation in vivo these complementary mechanisms for activating proteolysis, by mTOR inhibition and FoxO activation, function sequentially to enable cells to mobilize essential amino acids for the synthesis of proteins necessary for cell survival or energy production. Conversely, when nutrients and growth factors are ample, mTOR is activated, and FoxOs inactivated, the coordinate reduction in both proteolytic systems synergizes with the enhancement of protein synthesis to promote proteins accumulation. Although these three anabolic actions of 15796 | www.pnas.org/cgi/doi/10.1073/pnas.1521919112

mTORC1 are simultaneous and linked, they involve distinct downstream mechanisms, because mTORC1’s suppression of the UPS is independent of its targets that control translation or autophagy. This rapid activation of the UPS may also synergize with the activation of autophagy in helping eliminate abnormal toxic proteins and thus contribute to the beneficial effects of rapamycin in models of neurodegenerative disease (12) and in extending lifespan (11). Methods Cell Culture and Materials. HEK293, MEFs, and C2C12 myoblasts were maintained in DMEM with 10% (vol/vol) FBS in 5% (vol/vol) CO2 at 37 °C. C2C12 myotubes were obtained by incubating C2C12 myoblasts with differentiation medium [DMEM with 2% (vol/vol) horse serum] for 6 d in 8% (vol/vol) CO2 at 37 °C. Rapamycin and Akti1/2 were purchased from Calbiochem, concanamycin A from Santa Cruz, and HA-Ub-vinyl sulfone from Boston Biochem. All mouse experiments were performed with the approval of the Harvard Medical School Institutional Animal Care and Use Committee and are in accordance with the Guide for the Care and Use of Laboratory Animals (48). Determination of Total, Proteasomal, and Lysosomal Protein Degradation. This approach was described in detail and validated previously (14). Briefly, cells were incubated with 3H-Phe (5 μCi/mL) in standard culture medium for 20 h to label long-lived cell proteins, and then switched to a chase medium for 2 h that contains 2 mM nonradioactive Phe to prevent reincorporation of released radioactive amino acids. The medium was replaced with fresh chase medium containing the vehicle, BTZ (1 μM), or CCA (100-200 nM). One hour later, the mTOR inhibitors were added for another 1 h. Then multiple samples of the medium were collected at different times (up to 2 h) and mixed with trichloroacetic acid [TCA, final 10% (vol/vol)] to precipitate proteins. The TCA-soluble radioactivity in the medium at different times reflects the amount of prelabeled, long-lived proteins degraded and was expressed relative to the total radioactivity initially incorporated into protein. We generally used CCA-sensitive proteolysis to evaluate lysosomal degradation and CCA-resistant proteolysis to evaluate proteasome-mediated protein breakdown. All measurements were performed at least in triplicate, and the calculated rates of proteolysis are shown as “mean ± SEM.” P values were determined by two-tailed Student’s t test. Western Blotting. Cell proteins were extracted in lysis buffer [1% Triton X-100, 10 mM Tris pH 7.6, 150 mM NaCl, 30 mM Na pyrophosphate, 50 mM NaF, 5 mM EDTA, 0.1 mM Na3VO4 and protease inhibitor mixture (Roche)]. When measuring the levels of Ub conjugates, N-ethylmaleimide (10 mM) was added to the lysis buffer to inactivate most DUBs. Thirty micrograms of total proteins were separated by SDS/PAGE, transferred to PVDF membranes, and analyzed by Western blot using the ECL method (Amersham). We used antibodies against Akt, P-S6K (Thr389), P-Akt (ser473), P-Pras40 (T246), β-actin, α-tubulin, and Rpn11 from Cell Signaling; HMGCS1, P4D1, FL-76, and α-taxilin from Santa Cruz; Rictor and SUPT6H from Bethyl; FK2, Rpt5, α-subunits from Biomol. Western blot results were quantified by using ImageJ. Isolation and Mass Spectrometry Analysis of Ubiquitinated Proteins. Twelve 150-mm dishes of HEK293 cells were pretreated with cycloheximide for 1 h and then with either vehicle or Torin1 for another 1 h. After cell lysis in the presence of 10 mM N-ethylmaleimide to block deubiquitination, 800 μg of GST-UBA proteins were added to the lysates and incubated for 4 h and then with glutathione-resin for another 2 h. The immobilized proteins were extensively washed with PBS, digested with PreScission proteases overnight to release the ubiquitinated proteins together with the UBA domains from the GST-resin. These soluble proteins were then separated by SDS/PAGE, digested with trypsin, and analyzed by nanoscale-microcapillary reversed-phase liquid chromatography tandem mass spectrometry (LC-MS/MS) (49). ACKNOWLEDGMENTS. We thank Y. K. Ohsumi (Tokyo Institute Technology) for the generous gift of Atg5-deficient mouse embryonic fibroblasts; N. Sonenberg (McGill University) for the generous gift of 4E-BP1/2-deficient mouse embryonic fibroblasts; G. Thomas (University Cincinnati) for the generous gift of S6K1/2-deficient mouse embryonic fibroblasts; D. M. Sabatini (Massachusetts Institute of Technology) for the generous gift of Rictordeficient mouse embryonic fibroblasts; N. S. Gray (Harvard) and D. M. Sabatini (Massachusetts Institute of Technology) for the generous gift of Torin1; S. Thomas (Harvard) for plasmids for Ulks; and Takeda Pharmaceuticals for bortezomib and the Ube1 inhibitor. This work was supported by National Institute of Health Grants 5R01AR055255 and 5R01GM051923 (to A.L.G.); and grants from the Muscular Dystrophy Association of America, Packard Center, and ALS Association (all to A.L.G.).

Zhao et al.

Zhao et al.

PNAS | December 29, 2015 | vol. 112 | no. 52 | 15797

INAUGURAL ARTICLE

25. Hjerpe R, et al. (2009) Efficient protection and isolation of ubiquitylated proteins using tandem ubiquitin-binding entities. EMBO Rep 10(11):1250–1258. 26. Goldstein JL, Brown MS (1990) Regulation of the mevalonate pathway. Nature 343(6257):425–430. 27. Peterson TR, et al. (2011) mTOR complex 1 regulates lipin 1 localization to control the SREBP pathway. Cell 146(3):408–420. 28. Porstmann T, et al. (2008) SREBP activity is regulated by mTORC1 and contributes to Akt-dependent cell growth. Cell Metab 8(3):224–236. 29. Mashidori T, Shirataki H, Kamai T, Nakamura F, Yoshida K (2011) Increased alphataxilin protein expression is associated with the metastatic and invasive potential of renal cell cancer. Biomed Res 32(2):103–110. 30. Ohtomo N, et al. (2010) Expression of α-taxilin in hepatocellular carcinoma correlates with growth activity and malignant potential of the tumor. Int J Oncol 37(6): 1417–1423. 31. Hu X, et al. (2009) Genetic alterations and oncogenic pathways associated with breast cancer subtypes. Mol Cancer Res 7(4):511–522. 32. Iizuka M, et al. (2009) Histone acetyltransferase Hbo1: Catalytic activity, cellular abundance, and links to primary cancers. Gene 436(1-2):108–114. 33. Horii Y, et al. (2014) Expression of α-taxilin in the murine gastrointestinal tract: Potential implication in cell proliferation. Histochem Cell Biol 141(2):165–180. 34. Sakakibara S, et al. (2008) Developmental and spatial expression pattern of alphataxilin in the rat central nervous system. J Comp Neurol 511(1):65–80. 35. Schwanhäusser B, et al. (2011) Global quantification of mammalian gene expression control. Nature 473(7347):337–342. 36. Asano S, et al. (2015) Proteasomes. A molecular census of 26S proteasomes in intact neurons. Science 347(6220):439–442. 37. Tsvetkov P, et al. (2015) Compromising the 19S proteasome complex protects cells from reduced flux through the proteasome. eLife 4:e08467. 38. Renna M, et al. (2013) IGF-1 receptor antagonism inhibits autophagy. Hum Mol Genet 22(22):4528–4544. 39. Hsu PP, et al. (2011) The mTOR-regulated phosphoproteome reveals a mechanism of mTORC1-mediated inhibition of growth factor signaling. Science 332(6035): 1317–1322. 40. Yu Y, et al. (2011) Phosphoproteomic analysis identifies Grb10 as an mTORC1 substrate that negatively regulates insulin signaling. Science 332(6035):1322–1326. 41. Soucy TA, et al. (2009) An inhibitor of NEDD8-activating enzyme as a new approach to treat cancer. Nature 458(7239):732–736. 42. Khaminets A, Behl C, Dikic I (2015) Ubiquitin-dependent and independent signals in selective autophagy. Trends Cell Biol, 10.1016/j.tcb.2015.08.010. 43. Cohen S, Nathan JA, Goldberg AL (2015) Muscle wasting in disease: Molecular mechanisms and promising therapies. Nat Rev Drug Discov 14(1):58–74. 44. Lokireddy S, Kukushkin N, Goldberg AL (2015) cAMP-induced phosphorylation of 26S proteasomes on Rpn6/PSMD11 enhances their activity and the degradation of misfolded proteins. Proc Natl Acad Sci USA 112:E7176–E7185. 45. Myeku N, et al. (2015) 26S proteasome dysfunction and cognitive impairment caused by aggregated tau accumulation can be attenuated by PKA-mediated phosphorylation of proteasomes. Nat Med, 10.1038/nm.4011. 46. Kageyama S, et al. (2014) Proteasome dysfunction activates autophagy and the Keap1-Nrf2 pathway. J Biol Chem 289(36):24944–24955. 47. Mammucari C, et al. (2007) FoxO3 controls autophagy in skeletal muscle in vivo. Cell Metab 6(6):458–471. 48. National Research Council (2011) Guide for the Care and Use of Laboratory Animals (The National Academies Press, Washington, DC), 8th Ed. 49. Cohen S, Zhai B, Gygi SP, Goldberg AL (2012) Ubiquitylation by Trim32 causes coupled loss of desmin, Z-bands, and thin filaments in muscle atrophy. J Cell Biol 198(4): 575–589.

BIOCHEMISTRY

1. Goldberg AL, St John AC (1976) Intracellular protein degradation in mammalian and bacterial cells: Part 2. Annu Rev Biochem 45:747–803. 2. Ma XM, Blenis J (2009) Molecular mechanisms of mTOR-mediated translational control. Nat Rev Mol Cell Biol 10(5):307–318. 3. Feng Y, Yao Z, Klionsky DJ (2015) How to control self-digestion: Transcriptional, posttranscriptional, and post-translational regulation of autophagy. Trends Cell Biol 25(6): 354–363. 4. Chantranupong L, Wolfson RL, Sabatini DM (2015) Nutrient-sensing mechanisms across evolution. Cell 161(1):67–83. 5. Shimobayashi M, Hall MN (2014) Making new contacts: The mTOR network in metabolism and signalling crosstalk. Nat Rev Mol Cell Biol 15(3):155–162. 6. Dibble CC, Cantley LC (2015) Regulation of mTORC1 by PI3K signaling. Trends Cell Biol 25(9):545–555. 7. Thoreen CC, et al. (2009) An ATP-competitive mammalian target of rapamycin inhibitor reveals rapamycin-resistant functions of mTORC1. J Biol Chem 284(12): 8023–8032. 8. Thoreen CC, et al. (2012) A unifying model for mTORC1-mediated regulation of mRNA translation. Nature 485(7396):109–113. 9. Guertin DA, Sabatini DM (2009) The pharmacology of mTOR inhibition. Sci Signal 2(67):pe24. 10. Thomson AW, Turnquist HR, Raimondi G (2009) Immunoregulatory functions of mTOR inhibition. Nat Rev Immunol 9(5):324–337. 11. Harrison DE, et al. (2009) Rapamycin fed late in life extends lifespan in genetically heterogeneous mice. Nature 460(7253):392–395. 12. Menzies FM, Fleming A, Rubinsztein DC (2015) Compromised autophagy and neurodegenerative diseases. Nat Rev Neurosci 16(6):345–357. 13. Rock KL, et al. (1994) Inhibitors of the proteasome block the degradation of most cell proteins and the generation of peptides presented on MHC class I molecules. Cell 78(5):761–771. 14. Zhao J, et al. (2007) FoxO3 coordinately activates protein degradation by the autophagic/lysosomal and proteasomal pathways in atrophying muscle cells. Cell Metab 6(6):472–483. 15. Glickman MH, Ciechanover A (2002) The ubiquitin-proteasome proteolytic pathway: Destruction for the sake of construction. Physiol Rev 82(2):373–428. 16. Goldberg AL (2007) Functions of the proteasome: From protein degradation and immune surveillance to cancer therapy. Biochem Soc Trans 35(Pt 1):12–17. 17. Zhang Y, et al. (2014) Coordinated regulation of protein synthesis and degradation by mTORC1. Nature 513(7518):440–443. 18. Zhao J, Garcia GA, Goldberg AL (2015) Control of proteasomal proteolysis by mTOR. Nature, 10.1038/nature16472. 19. Kisselev AF, Goldberg AL (2001) Proteasome inhibitors: From research tools to drug candidates. Chem Biol 8(8):739–758. 20. Düvel K, et al. (2010) Activation of a metabolic gene regulatory network downstream of mTOR complex 1. Mol Cell 39(2):171–183. 21. Wang BT, et al. (2011) The mammalian target of rapamycin regulates cholesterol biosynthetic gene expression and exhibits a rapamycin-resistant transcriptional profile. Proc Natl Acad Sci USA 108(37):15201–15206. 22. Kimball SR, Do AN, Kutzler L, Cavener DR, Jefferson LS (2008) Rapid turnover of the mTOR complex 1 (mTORC1) repressor REDD1 and activation of mTORC1 signaling following inhibition of protein synthesis. J Biol Chem 283(6):3465–3475. 23. Nathan JA, Kim HT, Ting L, Gygi SP, Goldberg AL (2013) Why do cellular proteins linked to K63-polyubiquitin chains not associate with proteasomes? EMBO J 32(4): 552–565. 24. Chen JJ, et al. (2011) Mechanistic studies of substrate-assisted inhibition of ubiquitinactivating enzyme by adenosine sulfamate analogues. J Biol Chem 286(47): 40867–40877.