Cell Reports

Resource DUB-Resistant Ubiquitin to Survey Ubiquitination Switches in Mammalian Cells Miklo´s Be´ke´s,1 Keiji Okamoto,2 Sarah B. Crist,1 Mathew J. Jones,1 Jessica R. Chapman,3 Bradley B. Brasher,4 Francesco D. Melandri,4 Beatrix M. Ueberheide,1,3 Eros Lazzerini Denchi,2 and Tony T. Huang1,* 1Department

of Biochemistry & Molecular Pharmacology, New York University School of Medicine, New York, NY 10016, USA of Molecular and Experimental Medicine, The Scripps Research Institute, La Jolla, CA 92037, USA 3Proteomics Resource Center, New York University School of Medicine, New York, NY 10016, USA 4Boston Biochem, Cambridge, MA 02139, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.celrep.2013.10.008 This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-No Derivative Works License, which permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited. 2Department

SUMMARY

The ubiquitin-modification status of proteins in cells is highly dynamic and maintained by specific ligation machineries (E3 ligases) that tag proteins with ubiquitin or by deubiquitinating enzymes (DUBs) that remove the ubiquitin tag. The development of tools that offset this balance is critical in characterizing signaling pathways that utilize such ubiquitination switches. Herein, we generated a DUB-resistant ubiquitin mutant that is recalcitrant to cleavage by various families of DUBs both in vitro and in mammalian cells. As a proof-of-principle experiment, ectopic expression of the uncleavable ubiquitin stabilized monoubiquitinated PCNA in the absence of DNA damage and also revealed a defect in the clearance of the DNA damage response at unprotected telomeres. Importantly, a proteomic survey using the uncleavable ubiquitin identified ubiquitinated substrates, validating the DUB-resistant ubiquitin expression system as a valuable tool for interrogating cell signaling pathways. INTRODUCTION The ubiquitination status of a target protein is achieved via a delicate balance between two opposing forces: ubiquitin E3 ligases and DUBs. It has been postulated that the majority of proteins in a cell are regulated and modified by ubiquitin at some point (Hershko and Ciechanover, 1998); however, it has proved difficult to demonstrate the ubiquitination status of these proteins, as many of the modifications only exist transiently in vivo, owing largely to the multitude of DUBs present in cells. There are nearly 100 DUBs encoded in the human genome, which comprise approximately one-fifth of all proteases (Rawlings et al., 2010). DUBs are divided into five subfamilies based on catalytic mechanism and the fold of the active site domain (Reyes-Turcu et al., 2009). The largest family are the ubiquitin-

specific proteases (USPs), followed by the ovarian tumor DUBs (OTUs), the ubiquitin C-terminal hydrolases (UCHs), the Josephin DUBs, and, last, the JAMM/MPN+ family member DUBs. Except for the JAMM/MPN+ family of DUBs, which are zinc-dependent metalloproteases, all other characterized DUBs are cysteine proteases. As the ubiquitin-modified status of a protein can fundamentally alter its properties and change its biological role, it is critical to capture the ubiquitinated state of a protein in order to investigate its biological function both in vitro and in vivo. The ubiquitin system has recently been exploited with the design of unique ubiquitin mutants that inhibit specific DUBs (Ernst et al., 2013; Zhang et al., 2013). Although the specificity of such mutants is remarkable, generating a novel mutant for each DUB has to begin de novo and is quite laborious, especially when the physiological substrates of many DUBs remain unknown. In this study, we designed and generated a DUBresistant ubiquitin to capture and identify transiently ubiquitinated DUB substrates. Building on previous work in the SUMO conjugation and deconjugation pathway (Be´ke´s et al., 2011), we have generated a ubiquitin mutant (UbL73P) that is pleiotropically resistant to cleavage by multiple DUB families. This uncleavable ubiquitin mutant is conjugated to protein substrates in mammalian cells and leads to ubiquitin-conjugate stabilization. Ectopic expression of the DUB-resistant ubiquitin mutant stabilized monoubiquitinated PCNA, leading to the aberrant recruitment of translesion synthesis (TLS) polymerases in the absence of DNA damage, mimicking the effect of USP1 loss. Additional studies with DUB-resistant ubiquitin revealed a ubiquitin switch in the clearance of the DNA damage response (DDR) at shelterin-deficient chromosomal ends and captured ubiquitin-stabilized substrates by mass spectrometry. Our work provides a framework to study deubiquitination-dependent events both in vitro and in mammalian cells through the generation and use of the DUB-resistant ubiquitin tool. RESULTS Ubiquitin-L73P Is a DUB-Resistant Ubiquitin Mutant To establish a ubiquitin mutant that would be resistant to cleavage by DUBs, we mutated Leu73 of ubiquitin to Pro. Leu73 is the

826 Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors

A

DUB cleavage site ubiquitin residue #

P4 73

P3 74

P2 75

P1 76

WT : Leu - Arg - Gly - Gly L73P : Pro - Arg - Gly - Gly

B

L73P

C

SMT3-linear-di-Ub

WT

-

-

+

USP2CD

DUB cleavage

USP2CD

40

SMT3-Ub-Ub

Ulp1 cleavage

30

SMT3-Ub *

20

10

Ub

Smt3

D

di-ubiquitin

WT

L73P

-

E

M1-linked di-Ub

-

ubiquitination RXN with E1 + Ubc13 + mono-Ub

USP2CD

WT

L73P

ubiquitin

L73A

USP2CD

-

40

-

-

USP2CD 2

Ub 15 30 10

Ub1 Ub-Ub

15

10

USP2CD

40

Ub

F

G WT

WT

L73P

-

L73P

-

-

untreated

USP1/UAF1

-

Ub4 Ub3

36

14

Ub2

21

6

Ub1

31

USP2CD

USP5FL

USP7FL

USP8FL

untreated

WT L73P WT L73P WT L73P WT L73P WT L73P

USP2CD

USP5FL

USP7FL

USP8

WT L73P WT L73P WT L73P WT L73P WT L73P Ub 4

31

Ub 3

USP1 UAF1

J 2 SP

-

C

D

D C

2 SP U

D C

2

ba

tu

-

O

SP U

-

in

-1 in

D

2

C

tu

SP

O

U

-

AM

K63-linked-Ub4

-1

I ba

(+ ST AM

n1

SH

D C

2

tu ba i

O

SP

-

U

in

SH

AM

2

C

tu

SP

O

U

-

ba

D

-1

(+ ST AM

1)

H

K48-linked-Ub4

K48-linked-Ub4

1)

K63-linked-Ub4

Ub 1

6

U

188

Ub 2

14

USP2CD

STAM1

USP2CD

Otubain-1 AMSH USP2CD Otubain-1

31

4

Ub

2

Ub

2

Ub

2

Ub

1

Ub

6

WT

L73P

K63-linked Ub4

Ub1 WT

L73P

K48-linked di-Ub

Ub1 WT

L73P

K11-linked di-Ub

(legend on next page)

Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors 827

P4 position of the DUB cleavage site in the C terminus of ubiquitin (Figure 1A); the analogous mutation in SUMO2 (Figure S1A) results in a conjugatable but deconjugation-resistant SUMO (Be´ke´s et al., 2011). To test the ‘‘uncleavability’’ of UbL73P in the context of a linear peptide bond, we expressed recombinant linear diubiquitin (M1 linked) containing the L73P mutation in both ubiquitin moieties with an N-terminal Smt3-tag (Figure 1B) and tested it as a substrate for USP2CD (Figures 1C and S1B). Although the wild-type (WT) fusion protein is cleaved by USP2CD, the mutant (L73P) is not. To ensure that the Smt3-tag did not interfere with cleavage of the L73P di-Ub, the tag was removed via cleavage with Ulp1 and the di-Ub was purified to homogeneity and subjected again to USP2CD cleavage (Figure 1D). These results show that, in the context of a linear peptide bond, L73P is refractory to cleavage. To test whether UbL73P in the context of an isopeptide bond is also resistant to cleavage, we generated K63-linked ubiquitin chains using Ubc13 (UBE2N-Uev1a) in an in vitro ubiquitination reaction (Figure S1C, lanes 1–2 and 5–6). Whereas wild-type diubiquitin prepared using Ubc13 is cleaved by USP2CD (Figure 1E, lanes 1–4), di-UbL73P is completely resistant to cleavage (Figure 1E, lanes 5–8). Additionally, higher molecular weight, unanchored polyubiquitin chains, also prepared using Ubc13, are likewise resistant to cleavage in the context of UbL73P (Figure S1C, lanes 3–4 and 7–8). Interestingly, the more conservative L73A mutation on ubiquitin is only partially resistant to cleavage by USP2CD (Figure 1E, lanes 9–12). This suggests that it is the combination of the altered topology of the proline residue; the loss of the hydrophobic interaction provided by the leucine side chain; and the loss of its hydrogen-bonding ability to Asp295 of USP2 (Renatus et al., 2006) that renders UbL73P ‘‘uncleavable’’ (Figure S1D). Consistent with the latter being most significant, mutation of USP7 Asp295 to Ala results in an inactive enzyme (Hu et al., 2002). We show that purified linkage-specific ubiquitin chains produced in vitro are also resistant to cleavage by multiple USP family members (Figures 1F and 1G), by the K63-specific JAMM family member AMSH (Figure 1H) and by the K48-specific OTU-domain family member Otubain-1 (Figure 1I). Finally, we show that K11 linkages are also resistant to cleavage (Figure 1J). Collectively, these in vitro studies establish UbL73P as a pan-DUB resistant ubiquitin mutant, encompassing both cysteine protease and metalloenzyme DUBs.

Cleavage Resistant UbL73P Is Conjugated to Substrates Both In Vitro and In Vivo We have previously shown that UbL73P supports unanchored ubiquitin chain formation by Ubc13 as the E2 enzyme (Figures 1E and S1C). However, owing to the unique topology of UbL73P, we postulated that not all ubiquitin conjugation pathways will utilize it equally. Thus, we asked whether the two human E1 enzymes, Ube1 and Uba6, which have been shown to dictate downstream ubiquitination events in vivo (Jin et al., 2007), showed preference between wild-type and UbL73P. In an in vitro E1 charging reaction, Ube1 and Uba6 differ only slightly in UbL73P charging (Figure 2A); however, Uba6 cannot use UbL73P in charging UbcH7 in an E2-charging reaction (Figure 2B). Importantly, when UbL73P is utilized by Ube1 to charge an E2 enzyme, it depends on the active site cysteine of the E2, indicating that the charging is enzyme catalyzed (Figure S2A). Together, these results suggest that UbL73P usage in vivo is dictated at the E1 level; differences in E1 thioester formation between Ube1 and Uba6 could be due to the different residues that contact Leu73 in Ube1 and Uba6 (Figures S2B and S2C) during E1 charging (Olsen and Lima, 2013). On the other hand, there is no crystal structure available for a charged E2 enzyme still bound to an E1; therefore, the differential E2 charging ability of Uba6 in the context of UbL73P is difficult to explain mechanistically at this stage. Because Leu73 is part of the Ile36 hydrophobic patch of ubiquitin that is important in ubiquitin chain assembly (Komander and Rape, 2012), we asked whether the L73P mutation affects conjugation to substrates using E3 ligases. The recently solved crystal structures of RING-E2:E3 complexes highlight an intermediate role of Leu73 in E2-E3 interactions (Dou et al., 2012; Plechanovova´ et al., 2012). Consistent with the structural analysis, UbL73P is conjugated to the model substrate p53 by both RING- and HECT-domain E3 ligases with reduced efficiency (Mdm2, Figure 2C, and E6AP, Figure 2D, respectively). Importantly, however, UbL73P-conjugated p53-Ubn remains uncleavable by USP2CD (Figure 2E). We next sought to determine whether UbL73P, when expressed in mammalian cells, is incorporated into ubiquitin conjugates. We generated N-terminal HA-tagged wild-type and L73P ubiquitin constructs ending in Gly76 to allow for immediate conjugation. When expressed in U2OS cells, UbL73P

Figure 1. UbL73P Is a Pan-DUB DUB-Resistant Ubiquitin Mutant In Vitro (A) Surface representation of ubiquitin (PDB: 1UBQ) detailing its C terminus. The DUB cleavage site (P4-P1, aa 72–76 of ubiquitin) is shown in sticks (in black), with Leu73 shown in green. The Ile44 hydrophobic patch is colored brown on the surface, whereas the Ile36 hydrophobic patch is colored pink. The image was generated using PyMol. (B) Schematic representation of Smt3-tagged linear diubiquitin. Ubiquitin L73P is not cleaved by DUBs neither as a peptide or an iso-peptide bond. (C and D) Wild-type and L73P Smt3-linear-di-Ub, before (C) and after (D) Ulp1-cleavage and final purification, was cleaved with a serial dilution of USP2CD. (E) Untagged monoubiquitins were used to make K63-linked diubiquitins using Ubc13/Uev1a and then subjected to cleavage by USP2CD. (F) Recombinant USP1/UAF1 complex was used to cleave wild-type and L73P (NH) tetraubiquitin chains in vitro. (G) Linkage-specific tetra-ubiquitin chains (K48-linked, left panel, or K63-linked, right panel) made using either wild-type or L73P ubiquitin (nonhydrolyzable [NH] chains) were cleaved with an assortment of USP-family DUBs. (H) DUBs representing the USP family (USP2CD), the OTU-domain family (Otubain-1) and the JAMM-metallo-DUBs (AMSH) were used cleave wild-type and NH K63-linked tetraubiquitin chains. Otubain-1 serves as a negative control for K63. (I) Otubain-1 and USP2CD were used to cleave K48-linked diubiquitin. (J) USP2CD was used to cleave K11-linked diubiquitin. Dotted lines indicate cropping within the same gel. Assays were carried out at 37 C for 1 hr and analyzed by SDS-PAGE and Coomassie staining. See also Figure S1.

828 Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors

A

B GST-Ube1 -

WT

GST-Uba6

L73P

WT

5’ 15’ 5’ 15’

-

human E1

L73P

5’ 15’ 5’ 15’

GST-Ube1

-

ubiquitin

-

-

RXN @ 37ºC

-

160

WT

GST-Uba6

L73P

WT

5’ 15’ 5’ 15’

L73P

5’ 15’ 5’ 15’

human E1 -

ubiquitin

-

RXN @ 37ºC

E1~Ub E1 160

E1~Ub E1

E1 charging RXN

C p53 ubiquitination with Mdm2 -

WT

0’

10’

L73P 10’

30’

His-Ub

control

90’

RXN @ 37ºC

10’

p53-Ubn

80

30

UbcH7~Ub

p53-Ub1 20

p53

UbcH7

50

E2 charging RXN with UbcH7

D

E -

control

+

WT

-

WT L73P His-Ub + + p53 ubiquitination with E6AP p53-Ubn

80

80

+ -

L73P

+ +

-

+ -

Ub

+ +

Ub RXN with Mdm2 USP2CD cleavage post RXN p53-Ubn p53-Ub1

p53-Ub1 p53

50

50

p53 WB: anti-p53

38

USP2CD Ponceau S

F HA-ubiquitin

HA-ubiquitin

-

-

WT

L73P

188

L73P

49

98

62

WT

62

PCNA-Ub 1

HMW Ubn

38

PCNA 28

49

WB: PCNA (short exposure)

62

PCNA-Ub 2

38 49

PCNA-Ub 1

28

Ub2

38

PCNA 28

14

Ub1

WB: PCNA (long exposure)

98

6

MCM6

Figure 2. UbL73P Is Conjugated to Substrates In Vitro and In Vivo (A) E1 charging assay with GST-tagged human Ube1 and Uba6 using wild-type and L73P ubiquitin. (B) E2 charging assay using human GST-tagged Ube1 and Uba6 as the E1 and human UbcH7 as the E2, using wild-type and L73P ubiquitin (F). E1 and E2 charging reactions were carried out at 37 C in the absence of DTT and analyzed by nonreducing SDS-PAGE and SYPRO staining. (C and D) p53-ubiquitination reactions with wild-type and L73P ubiquitin using either a RING-finger E3 ligase, Mdm2 (C), or a HECT-domain E3 ligase, E6AP (D). Ubiquitination reactions with the E3 ligases were analyzed by denaturing SDS-PAGE and western blotting with a p53 antibody. Dotted lines indicate cropped images from the same gel. (E) Mdm2-generated ubiquitin chains on p53 using WT or L73P ubiquitin were cleaved with USP2CD following ubiquitination reactions. (F) Expression of HA-tagged wild-type and L73P ubiquitin in U2OS cells results in stabilization of Ub-conjugates by L73P, such as PCNA-Ub. See also Figure S2.

Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors 829

A

B

C

E

D

F

(legend on next page)

830 Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors

forms Ub conjugates similar to wild-type Ub, with a modest increase in higher -molecular weight (HMW) Ub conjugates (Figure 2F, left panel), likely forming heterogenous chains with endogenous ubiquitin. Under these conditions, levels of ectopically expressed ubiquitin do not reach endogenous levels, as judged by total Ub (Figures S3A and S3B). Intriguingly, expression of UbL73P resulted in a marked increase in the monoubiquitination of PCNA (Figure 2F, right panel), a substrate previously shown to be monoubiquitinated under DNA damage conditions (Hoege et al., 2002). Nonspecific ubiquitination was not observed for other proteins, such as MCM6 (Figure 2F, bottom). Importantly, we show that PCNA modification with cleavage-resistant modifiers is specific to ubiquitin, as cleavage-resistant SUMO2 does not stabilize PCNA conjugates under normal conditions (Figure S3C). The monoubiquitination of PCNA is usually reversed by USP1 (Huang et al., 2006); however, USP1 does not cleave UbL73P chains (Figure 1F). These results strongly suggest that the UbL73Pstabilized PCNA monoubiquitination is a result of the failure of USP1 to cleave UbL73P from PCNA in cells. UbL73P Expression Inhibits the Deubiquitination of PCNA in the Absence of DNA Damage We next focused on how stabilized UbL73P conjugates lead to higher levels of PCNA monoubiquitination (Figure 2F). UbL73P conjugates produced from cells remain refractory to an in vitro DUB (USP2) cleavage assay (Figure 3A, lanes 1–8). Interestingly, whereas expression of both UbL73A and UbG76A give rise to similar patterns of HMW ubiquitin conjugates in cells, they remain relatively sensitive to cleavage by USP2CD in vitro (Figure 3A, lanes 9–16). An earlier study showed that UbG76A can accumulate as K48-linked unanchored ubiquitin oligomers in cells (Hodgins et al., 1992). Collectively, this suggests that UbG76A is not universally refractory to cleavage by DUBs, but it is resistant only to USP5 (IsoT)-mediated deubiquitination (Dayal et al., 2009), which requires an intact C-terminal GlyGly motif for unanchored chain deconjugation (Wilkinson et al., 1995). In agreement with the DUB cleavage sensitivity of different ubiquitin mutants in vitro (Figures 1E and 3A), expression of neither UbL73A nor UbG76A stabilize PCNA monoubiquitination in cells (Figure 3B). Importantly, we show that PCNA monoubiquitination by UbL73P is restricted to the canonical ubiquitination lysine of PCNA, because coexpression of a K164R

PCNA mutant does not support monoubiquitination with UbL73P (Figure 3C). We also show that UbL73P-stabilized PCNA monoubiquitination is dependent on the endogenous E3 ligase for PCNA, Rad18 (Kannouche et al., 2004), because small interfering RNA (siRNA) knockdown of Rad18 reduces PCNA monoubiquitination levels (Figure 3D). These results strongly suggest that UbL73P conjugation is specific to a physiological substrate and its lysine residue, and it occurs via its endogenous E3 ligase. Finally, we show that UbL73P-stabilized PCNA monoubiquitination aberrantly recruits TLS polymerase ETA to the replication fork, represented by nuclear foci (Figure 3E). This is in accordance with earlier work showing that knockdown of USP1, which leads to the accumulation of monoubiquitinated PCNA even in the absence of DNA damage, is sufficient to recruit TLS polymerases to the replication fork and forms nuclear foci (Jones et al., 2012). These results demonstrate that UbL73P is utilized similarly as wild-type ubiquitin and is conjugated onto protein substrates in UbL73P-expressing cells. This ‘‘proof-ofprinciple’’ highlights the use of UbL73P as a tool to reveal ubiquitin switches (as demonstrated by the balance between ubiquitin conjugation and deconjugation cycles), such as for PCNA in unperturbed cells (Figure 3F). UbL73P Expression Attenuates the Clearance of 53BP1 Foci after Telomeric DDR in Shelterin-Deficient MEFs We next sought to determine whether UbL73P can be used to interrogate DNA repair pathways that are regulated by ubiquitin switches. Distinct E3 ligases and DUBs are critical regulators of the double-strand break repair and response in mammalian cells (Jackson and Durocher, 2013). Signaling and repair emanating from DNA damage recognition is carried out, in part, through mono- and polyubiquitin interactions with different ubiquitinbinding domains (UBDs) on specific DNA damage response (DDR) proteins. To determine if UbL73P can be utilized and recognized by DDR proteins, we first addressed whether UbL73P conjugates are recognized by UBDs other than those of the TLS polymerases. Indeed, UBD pull-down experiments using the UIMs (ubiquitin-interacting motifs) of RAP80 (Figure 4A) and S5a (Figure S4A) indicate that polyubiquitin conjugates from cell extracts that have incorporated UbL73P are efficiently captured by different UBDs. To address whether K63 linkagespecific polyubiquitin chains can be recognized and captured by the RAP80 UBD in the context of the UbL73P-stabilized chains, UbL73P containing only a single lysine site at K63 was

Figure 3. Dynamic PCNA Monoubiquitination Revealed by UbL73P (A) HA-Ub constructs were expressed in HEK293 cells, lysed under nondenaturing conditions, and the lysates were treated with a serial dilution of USP2CD. (B) HA-Ub constructs were expressed in U2OS cells and lysed under denaturing conditions. (C) PCNA with a K164R mutation does not support monoubiquitination by UbL73P. HA-PCNA and myc-ubiquitin constructs were coexpressed in U2OS cells and lysed under denaturing conditions. (D) UbL73P is conjugated in a Rad18-dependent manner in vivo. U2OS cells were treated with siRNAs for 24 hr and then transfected with HA-ubiquitin for another 48 hr when cell lysates were prepared under denaturing conditions. (E) UbL73P-stabilized PCNA-Ub recruits polymerase eta in the absence of DNA damage. HA-tagged Ubs were coexpressed with GFP-tagged Pol eta in U2OS cells, fixed in methanol, mounted with DAPI, and analyzed by immunofluorescence microscopy using GFP antibodies for nuclear GFP-ETA foci. The inlets show representative images and western blots for the cell lysates analyzed. Error bars represent SD of the mean, n = 3. (F) Dynamic PCNA monoubiquitination revealed by UbL73P. Under normal circumstance in S-phase (top panel) the E3 ligase Rad18 monoubiquitinates PCNA, which is counterbalanced by USP1, the DUB for PCNA. When a conjugatable, but not deconjugatable ubiquitin (L73P, bottom panel) is expressed in cells, the USP1-arm of the dynamic PCNA monoubiquitation cycle is poisoned, leading to the stabilization of PCNA-Ub. See also Figure S3.

Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors 831

A input (5%) -

B

RAP80-UIM pull-down

WT L73P -

L73P K48- K63only only

WT L73P HA-Ub

L73P

HA-Ub

K48- K63- available Lys only only

188 98

HMW-Ub

188

n

98

62

HMW-Ub

62

49

49

38

38

28

28

Ub 3 Ub 2

14

14

6

6

WB: HA

WB: HA

98

input (5%)

MCM4

RAP80-UIM pull-down

D

C TRF2F/F; R26-CreER + OHT HA-Ub Day 4

HA-Ub L73P

Day 7

Day 4

Day 7

53BP1

Day 4 Day 4

Merge

40 30 20 10 0

Day 7

50

Day 7

% of cells with 53BP1 foci

TTAGGG

% of cells with 53BP1 foci

50

40 30 20 10 0

30 HA-Ub-WT

30 HA-Ub-L73P

E HA-Ub+ OHT

HA-Ub L73P + OHT

72.6% TRF2F/F; R26-CreER

80

% of fused chromosomes

73.2%

70 60

- OHT

50

+ OHT

40 30 20 10 0

Vector

HA-Ub WT

HA-Ub L73P

(legend on next page)

832 Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors

expressed in cells and was shown to form polyubiquitin chains that could be selectively enriched by the RAP80 UBD in a pulldown experiment (Figure 4B). A critical DDR regulator is the RAP80 DUB complex, containing the K63 linkage-specific metalloprotease BRCC3 (Kim et al., 2007; Sobhian et al., 2007; Wang et al., 2007). Based on the observation that RAP80 UIMs bind UbL73P-stabilized K63-Ub conjugates and the recent report supporting the requirement of BRCC3 to prevent the accumulation of 53BP1 foci at telomeres in the absence of the shelterin component TRF2 (Okamoto et al., 2013), we sought to determine whether expression of UbL73P contributes to the dynamics of 53BP1 foci accumulation. In mouse embryonic fibroblasts (MEFs) lacking TRF2, spontaneous activation of DDR occurs, leading to gH2AX (Ser139 phosphorylation) and 53BP1 foci accumulation at the chromosome ends. These DNA damage foci are cleared following the nonhomologous end-joining (NHEJ)mediated ‘‘repair’’ of dysfunctional telomeres that results in end-to-end chromosome fusions (Celli and de Lange, 2005). To determine whether expression of UbL73P prevents dissipation of the DNA damage signal in this experimental setting, we compared lentiviral transduction of UbWT or UbL73P in 4-hydroytamoxifen-treated Trf2Floxed/Floxed MEFs carrying an inducible Cre recombinase (Rosa26-CreERT2) to activate DDR (Okamoto et al., 2013). Following Cre activation, we monitored 53BP1 localization at telomeres immediately after TRF2 depletion (day 3) and at a later time point when the vast majority of telomeres have been processed by the NHEJ-mediated repair (day 7). Intriguingly, whereas the initial accumulation of 53BP1 foci is not affected by UbL73P, the clearance of 53BP1 foci is significantly attenuated at day 7 (Figures 4C and 4D, right panels), suggesting that deubiquitination (likely through BRCC3) is required for 53BP1 clearance. Importantly, UbL73P expression does not globally activate DDR on its own (as observed by the absence of spontaneous gH2AX levels) nor does it affect the rate of NHEJ at telomeres, as judged by occurrence of telomere fusions (Figure 4E). On the contrary, expression of UbL73P does not affect the clearance of DNA damage foci at DSBs occurring randomly in the genome, as shown by the dissipation of 53BP1 foci in cells treated with ionizing radiation (IR) or bleomycin (Figures S4B– S4D). These results suggest that although clearance of DDR factors at TRF2-depleted telomeres depends strictly on a UbL73P-stabilized protein substrate(s), additional pathways may act at DSBs that occurred randomly in the genome. Similarly, activation of the DDR at TRF2-depleted telomeres is strictly dependent on the ATM DDR kinase, whereas irradiation

induced foci (IRIFs) are largely unaffected in ATM-deficient cells (Denchi and de Lange, 2007). Identification of Ubiquitin-Stabilized Substrates in HeLa Cells Using an UbL73P Expression System We sought to establish the utility of UbL73P in identifying ubiquitin-stabilized substrates by creating a stable cell line expressing UbL73P, purifying Ub conjugates and identifying the substrates and ubiquitination sites by mass spectrometry. We generated a doxycyclin-inducible HeLa cell line stably expressing low levels of FLAG-Ub-WT (Flp-in-WT) and FLAGUb-L73P (Flp-in-L73P) (Figure 5A). Purifying FLAG conjugates from Flp-in cells confirmed PCNA-Ub stabilization using UbL73P (Figure 5B), whereas the low levels of FLAG-Ub-L73P expressed did not affect the NFkB signaling pathway, because ubiquitin-mediated proteasomal degradation of IkBa following TNF-a treatment was comparable in both Flp-in-WT and Flpin-L73P cells (Figure S4E). Next, we scaled up and purified FLAG-Ub conjugates from Flp-in-L73P cells and subjected them to mass spectrometry-based protein identification. We identified potential ubiquitinated substrates, including previously validated ones (Table S1). Importantly, we were also able to identify ubiquitination sites for a handful of proteins (Figure 5C), including Lys164 of PCNA (Figure 5D), validating our approach. Interestingly, we identified Ubc13 as a ubiquitinated substrate at Lys92 (Figure 5E). The ubiquitination stabilization of Ubc13 by UbL73P, and not by WT ubiquitin, was confirmed in U2OS cells for endogenous Ubc13 (Figure 5F) as well as by coexpression of FLAG-Ubc13 and HA-ubiquitins (Figure 5G). These results unequivocally demonstrate the utility of UbL73P in stabilizing and identifying ephemeral ubiquitin conjugates that would otherwise constantly undergo deubiquitination by DUBs. DISCUSSION In summary, we have introduced and characterized a ubiquitin point mutant capable of conjugating to cognate substrates and incorporating into ubiquitin chains, yet remains refractory to cleavage by DUBs. This provides a unique tool to enable the generation, identification, and study of substrates with stabilized ubiquitination states. Previous work by Be´ke´s et al. (2011) on the related ubiquitin-like molecule SUMO2 showed that mutation in the P4 position of the SUMO cleavage site for deSUMOylating enzymes (SENPs) from glutamine to proline (Figure S1A, Q90P, which is also found naturally in the pseudogene, SUMO4 [Bohren et al., 2007]), results in resistance to cleavage by deSUMOylating

Figure 4. Ubiquitin-Conjugate Stabilization Reveals a Telomeric DNA Damage Response Phenotype RAP80-mediated pull-down of UbL73P conjugates. (A and B) HA-ubiquitin constructs were expressed in HEK293 cells and lysed under nondenaturing conditions, and Ub conjugates were pulled down with RAP80-conjugated beads and then analyzed by SDS-PAGE and western blotting for the indicated antibodies. (C) TRF2F/F; Rosa26 CRE-ER MEFs infected with the indicated constructs were treated with 4-hydroxytamoxifen (+OHT) and stained for 53BP1 (red), telomere DNA (green), and DAPI (blue). (D) Quantification of cells with less than ten 53BP1 foci (30) per nucleus at the indicated time points following OHT treatment. (E) Metaphase spreads were harvested from TRF2F/F; Rosa26 CRE-ER MEFs 7 days following OHT treatment. Cells were infected with either HA-Ub or HA-UbL73P and stained for telomere DNA (green) and DAPI (red). Percentages of chromosomes with fusions are indicated. Bar graph in the lower panel indicates the percentage of chromosome fusions in MEFs infected with the indicated constructs and either treated with tamoxifen (+OHT) or untreated ( OHT). See also Figure S4.

Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors 833

A

B

HeLa T-REx Flp-In cells

WT FLAG-Ub-WT

FLAG-Ub-WT

FLAG-Ub-L73P

FLAG-Ub-L73P

0

24

L73P 48

0

24

Flp-In cells 48

hrs with DOX

FLAG IP

PCNA-Ub

1

38

PCNA 28

Flp-mediated recombination

WB: PCNA

Doxycycline Induction

FLAG IP

C #

Protein name

Coverage

GlyGly-modified lysine (KGG) peptide

XCorr

Experimentally validated

1

PCNA

77.01 %

DLSHIGDAVVISCAKGGDGVK

5.54

this study & Hoege C (2002) Nature

2

Ubc13

53.95 %

ICLDILKGGDK

2.23

this study

3

Ube2T

18.27 %

ICLDVLKGGLPPK

3.04

Machida YJ (2006) Mol Cell

4

Ubiquitin (K11)

60.90 %

TLTGKGGTITLEVEPSDTIENVK

5.05

established Ub-Ub site

5

Ubiquitin (K48)

60.90 %

LIFAGKGGQLEDGR

2.66

established Ub-Ub site

6

PCNA-associated factor (PAF15)

51.35 %

KGGVLGSSTSATNSTSVSSR

4.36

Povlsen LK (2012) NCB

7

Histone H2A

36.72 %

VTIAQGGVLPNIQAVLLPKGGK

4.91

Zhou W (2008) Mol Cell

8

HINT1

41.03 %

EIPAKGGIIFEDDR

2.88

not yet validated

9

Isoform 4 of Bcl-6 corepressor

1.47 %

LIVNKGGNAGETLLQR

3.40

not yet validated

10

Actin

44.53 %

EITALAPSTMKGGIK

3.55

not yet validated

11

Nedd8

25.93 %

LIYSGKGGQMNDEK

3.43

not yet validated

F

b4 b5 b6

b7 b8 b9

H

D A V V I

I

G

y17 y16 y15 y14 y13 y12 y11 y10

S C* A KGG D

y9

y6

y8

y5

G V K y3

Ubc13

[M+2H]2+ % Relative Abundance

Ubc13-Ub1

28

b4

100

y14

14

y17+2

WB: Ubc13

49

b5 y8

b9 y9 b8 b10 b7 y +2 16

y4

y3

y5 b6 y6

200

300

400

500

E

600

b11

700

800

900

m/z

1000

1100

y10

y11

1200

y15

PCNA

y13

1300

1400

WB: PCNA 1500

1600

1700

1800

b2

I

C*

USP5

98

L D

I

y6

y5

L KGG

D

y4

y2

y3

K

WB: USP5

a2

100

PCNA-Ub1

38

y12

G

y3

P

0

y4

P

L S

% Relative Abundance

b2

W T L7 3

D

W T L7 3

D

+ + +

HA-ubiquitin FLAG-Ubc13

38

Ubc13-Ub2

28

Ubc13-Ub1

y4

y2

Ubc13 y6

WB: FLAG

y5 0

200

300

400

500

600

m/z

700

800

900

1000

(legend on next page)

834 Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors

enzymes, allowing the generation of stabilized SUMO2conjugates in cells. We then asked whether mutation of the corresponding residue (Leu73) on ubiquitin could affect ubiquitin cleavage reactions by DUBs, and we indeed show that it does. Interestingly, in an alanine-scanning mutagenesis study of yeast ubiquitin more than a decade ago, it was shown that UbL73A allows for ubiquitin conjugation; however, the mutant is partially defective in an S. cerevisiae endocytosis assay (Sloper-Mould et al., 2001). A recent study has also identified bulky Leu73 mutations that differentially affect conjugation and deconjugation activities (Zhao et al., 2012). Furthermore, yeast provided with the UbL73A mutant as the only source of ubiquitin cannot support vegetative growth (Sloper-Mould et al., 2001). It is possible that the Leu73 mutant phenotype of ubiquitin in yeast is a composite of defects in conjugation as well as deconjugation. Leu73 contributes to part of the hydrophobic patch in ubiquitin centered around Ile36, which is utilized by some E3 ligases to synthesize polyubiquitin chains (Plechanovova´ et al., 2012). Therefore, in accordance with our in vitro E1/E2/E3 conjugation data, it is possible that certain substrates, whose E3 ligases solely rely on ubiquitin Ile36 hydrophobic interactions, would not be efficiently conjugated by UbL73P. This selectivity at the E3 level, together with the selectivity of the E1 enzymes, Ube1 and Uba6, in differentially charging E2 enzymes, suggests that the conjugation of UbL73P in vivo would be skewed toward conjugation pathways that can tolerate it. Nevertheless, those UbL73P conjugates would remain stable and resistant to DUBs in cells, as is the case for monoubiquitinated PCNA, the identified Ubc13, unknown factor(s) in the telomeric DDR pathway, and others. Increased levels of PCNA monoubiquitination by UbL73P expression in a damage-independent manner mimics the phenotype observed for USP1 knockdown (Huang et al., 2006; Jones et al., 2012). USP1 is the only DUB to date shown to remove ubiquitin from PCNA in vivo. This finding reveals the highly dynamic nature of PCNA monoubiquitination in undamaged cells by the Rad18-USP1 E3-DUB cycle and underscores the crucial regulatory role of USP1 in maintaining PCNA in an unubiquitinated state. This regulation ensures that PCNA monoubiquitination will not serve as a platform to recruit low-fidelity translesion synthesis (TLS) polymerases (Lehmann et al., 2007) in the absence of DNA damage. Failure to maintain appropriate PCNA monoubiquitination levels could result in increased TLS polymerase recruitment, such as that of poly-

merase kappa (Jones et al., 2012), which increases genomic instability. Despite the important role of ubiquitination-dependent mechanisms in DSB repair (Jackson and Durocher, 2013), UbL73P displayed no global defects in DNA repair mechanisms following bleomycin treatment or ionizing radiation, perhaps because UbL73P was incompletely utilized or because the need for deubiquitination was bypassed. However, the phenotype brought about by UbL73P, in the context of the activation of spontaneous DNA damage response in the absence of the shelterin complex at telomeres, suggests that UbL73P is a functional player and results in the Ub conjugate stabilization of one or more factors that are required for efficient 53BP1 clearance at chromosomal ends. This is in accordance with a recent finding showing that telomeric and genomic DDR pathways are quite different (Cesare et al., 2013): telomeric DDR does not activate checkpoint signaling and could rely on entirely different sets of effector proteins, some of which could be sensitive to stabilized ubiquitination. The telomeric UbL73P phenotype mimics the one observed with the knockdown of BRCC3, which opposes the activity of RNF168 at telomeres (Okamoto et al., 2013). The identity of the ubiquitinated substrate(s) that are responsible for the activation/regulation of telomeric DDR is currently unknown. In the future, it will be important to discern global versus telomeric ubiquitination substrates in DNA damage response pathways. In our exploratory UbL73P substrate identification survey, we validated the approach by identifying known ubiquitination sites on PCNA and others (Povlsen et al., 2012; Zhou et al., 2008). We also identified several ubiquitinated substrates stabilized by UbL73P. We were able to confirm the ubiquitination of Ubc13 on Lys92, a previously indicated site of ISG15 modification (Giannakopoulos et al., 2005). Additionally, we identified the site of ubiquitination on Ube2T, another E2 enzyme that has been previously shown to be ubiquitinated (Machida et al., 2006). Intriguingly, several other E2 Ub-conjugating enzymes were found in our mass spectrometry results. Whether reversible E2 ubiquitination is a regulatory posttranslational modification common to a variety of E2 enzymes remains to be determined. The utility of DUB-resistant ubiquitin tools is highly promising. First, it is possible to ubiquitinate substrates using UbL73P in vitro and assay the stable Ub conjugate in subsequent binding or affinity purification experiments in search of binding partners. Second, the UbL73P-stabilized ubiquitin chains of a defined linkage could also be utilized to identify substrates modified by a

Figure 5. UbL73P Reveals Stabilized Ubiquitination of Ubc13 and Other Targets (A) Schematics for the generation of HeLa Flp-in cells stably expressing ubiquitin constructs. (B) Anti-FLAG immune precipitation of Flp-in cells treated for the indicated times with doxycycline to induce transgene expression, probed with anti-PCNA antibody. (C) Selected list of proteins conjugated by UbL73P and the peptides containing the Ub-modified Lys residues. (D) MS/MS spectrum of the doubly charged ion of the peptide (DLSHIGDAVVISCAKGGDGVK carboxymethylated on the Cys (*) residue and carrying a GlyGly modification on the first lysine residue) for Lys164 of PCNA. (E) MS/MS spectrum of the triply charged ion of the peptide (ICLDILKGGDK carboxymethylated on the Cys (*) residue and carrying a GlyGly modification on the first lysine residue) for Lys92 of Ubc13. Observed peptide bond cleavages are indicated in the sequences. The corresponding theoretical N-terminal (b-type ion) and C-terminal (y-type ions) ion series for the observed fragment ions are shown above and below the sequence, respectively. Neutral loss of water from y- and b-type ions are not indicated in the spectra. (F) HA-Ub constructs were expressed in U2OS cells, lysed under nondenaturing conditions, and probed with the indicated antibodies. (G) FLAG-Ubc13 and HA-Ub constructs were coexpressed in U2OS cells, lysed under denaturing conditions, and probed with FLAG antibody. See also Figure S4 and Table S1.

Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors 835

particular ubiquitin chain linkage or identify binding proteins unique to a particular ubiquitin chain linkage. This approach will be most beneficial for chains utilizing hitherto uncharacterized linkages, such as Lys29 and Lys33, and will also allow the purification and identification of ubiquitin receptors that recognize these unique chains. Finally, as unanchored oligomeric ubiquitin chains appear to be stabilized by both UbG76A and by UbL73P, such DUB-resistant ubiquitin mutants have the potential to shed light on the nature and the dynamics of de novo unanchored ubiquitin chain formation, both in resting cells and in a signal inducible manner. This has been a relatively unexplored area of research, as only recently have unanchored ubiquitin chains been implicated in kinase activation (Xia et al., 2009) and in antiviral immunity (Zeng et al., 2010). As long as the conjugation competency for a particular pathway is adequate, we envision the use of DUB-resistant ubiquitins and other isopeptidase-resistant ubiquitin-like molecules by researchers who seek to interrogate ubiquitin and ubiquitin-like-mediated biological processes in a wide range of experimental settings. EXPERIMENTAL PROCEDURES In Vitro Ubiquitination and Deubiquitination Reactions Unless otherwise indicated, ubiquitination and E1- and E2-charging reactions were carried out as previously described (Petroski and Deshaies, 2005). Briefly, in 30 mM HEPES (pH 7.5), 100 mM NaCl, 5 mM MgCl2, and 20 mM ubiquitin (WT or its mutants) were incubated with E1 and E2 enzymes at 150 and 500 nM, respectively, in the presence or absence of 2 mM ATP for the indicated times at 37 C and boiled immediately in sample buffer (±5 mM DTT for nonreducing/reducing gels) to terminate the reaction. For the deubiquitination assay of Ubc13(Ubc13/Uev1a heterodimer)-conjugated diubiquitin, the ubiquitin assay described above was performed for 1 hr, and then the reaction was terminated with 25 mM EDTA for 10 min and diluted (4/5) into a serial dilution of USP2CD, and the mix was incubated for another hour at 37 C. For the single time point in vitro DUB assays, 2 mg wild-type of L73P (nonhydrolyzable, NH) ubiquitin tetramers were cleaved with 100 nM DUBs for 1 hr at 37 C and analyzed by SDS-PAGE and Coomassie staining. For USP2CD serial dilutions (with 1 mg of linear chains or with Ubc13-produced K63-chains or with 10 mg total cell lysates), USP2CD was prepared at a starting concentration of 1 mM with 1/10 serial dilution in assay buffer or in the cell lysate with a final concentration of 5 mM DTT. Ubiquitination reactions for p53 using E3 ligases were carried out with E3 kits from Boston Biochem according to the protocols described therein, and then, where indicated, reactions were quenched with 50 mM EDTA and further incubated with 1 mM USP2CD for 30 min at 37 C. Cell Culture, Plasmid Transfection, Lysate Preparation, and Immunoprecipitation U2OS and HEK293 cells were maintained as previously described (Huang et al., 2006). Transfections were carried out using Fugene (Promega) or Hiperfect (QIAGEN), for plasmids or siRNA oligos, respectively, according to the manufacturer’s protocol. The siRNA used to knock down Rad18 has been described previously (Huang et al., 2006). Cells were routinely harvested in PBS, and the pellets were frozen at 80 C prior to lysis. Cells were lysed in either denaturing SDS buffer (100 mM Tris [pH 6.8], 2% SDS and 20 mM b-Me) for direct analysis by SDS-PAGE or in nondenaturing buffer (50 mM Tris [pH 7.6], 150 mM NaCl, 2 mM EDTA, 0.5% NP-40, and protease inhibitor cocktail [Roche], and benzoase [Novagen]) for immunoprecipitations (IPs) and ubiquitin-receptor binding assays. Ubiquitin-Receptor Binding Assays Cell lysates prepared under nondenaturing conditions (150 mg) were incubated with 10 ml agarose-conjugated ubiquitin-receptors (RAP80-UIMs or

S5a/Angiocidin, Boston Biochem) for 1 hr at 4 C in binding buffer (50 mM Tris [pH 8.0], 150 mM NaCl, 0.1% NP-40, 5 mM DTT, supplemented with a protease inhibitor cocktail [Roche]) and then washed three times in binding buffer and analyzed by SDS-PAGE and western blotting. Plasmids and Recombinant Protein Expression For the purification of untagged human ubiquitins (linear diubiquitin and monoubiquitin mutants), ubiquitin (residues 1–76) was cloned as a His6SMT3-fusion protein (plasmid is gift from Chris Lima). Briefly, SMT3-Ub proteins were expressed in E. coli codon-plus cells at 30 C for 3 hr by inducing expression with 1 mM IPTG. His6-SMT3-Ub proteins were purified by Ni2+NTA and size exclusion chromatography on a Superdex-200 26/60 column in 20 mM Tris (pH 8.0), 350 mM NaCl, and 1 mM b-Me. The purified fusion protein was cleaved overnight at 4 C with Ulp1 at a molar ratio of 1:1,000; next, untagged ubiquitin was separated from His6-SMT3 by a final ion-exchange step on a MonoQ 10/10 column. Untagged ubiquitin proteins produced this way have additional SerGly residues at their N termini owing to the cloning strategy using BamHI-XhoI. For mammalian expression, ubiquitin (residues 2–76) was subcloned into pcDNA3 with an N-terminal HA tag using the restriction sites HindIII and XhoI and has been used as a template to obtain the indicated point mutants by site-directed mutagenesis. The catalytic domain of human USP2a (residues 259–605) was cloned with a C-terminal His-tag in pET21a and expressed as previously. Briefly, USP2CD was expressed in E. coli codon-plus cells at 30 C for 5 hr by inducing with 1 mM IPTG. USP2CDHis was then purified by Ni2+-NTA and size-exclusion chromatography on a Superdex-75 26/60 column in 20 mM HEPES (pH 7.0), 350 mM NaCl, and 1 mM b-Me, followed by a final ion-exchange step on a MonoS 10/10 column. Recombinant S. pombe E1 (Uba1) was a gift from Shaun K. Olsen; other enzymes have been purchased from Boston Biochem. USP1/UAF1 was purchased from Ubiquigent. All constructs have been verified by sequencing and all recombinant proteins have been aliquoted and stored at 80 C until use. Antibodies The following antibodies were used in this study: antiubiquitin (P4D1, sc-8017, Santa Cruz Biotechnology), anti-His (mouse-monoclonal, GenScript), anti-HA (Mono HA.11, MMS-101R, Covance), anti-PCNA (PC10, sc-56, Santa Cruz), anti-MCM6 (A300-194A-2, Bethyl Laboratories), anti-Rad18 (A301-340A, Bethyl Labs), anti-MCM4 (A300-193A, Bethyl Labs), anti-myc (c-3956, Sigma), anti-GFP (B-2, sc-9996, Santa Cruz), and anti-p53 (Boston Biochem). Immunofluorescence Microscopy Cells were grown and processed in 8-well Lab-Teks II ChamberSlide System slides from Nunc. For the bleomycin recovery experiment, cells were treated with 5 mg/ml bleomycin sulfate (Calbiochem) for 1 hr at 37 C and then, where indicated, recovered in full media for 24 hr and fixed using methanol fixation and stained with HA and 53BP1 antibodies. GFP-Pol-eta nuclear foci formation was detected after methanol fixation, and cells were stained with anti-GFP for 2 hr at room temperature. The secondary antibodies used were Alexa-conjugated goat anti-mouse and anti-rabbit. Slides were mounted in Vectashield with DAPI (Vector Laboratories). Images were deconvolved using Softworx software (Applied Precision). The images were opened and then sized and placed into figures using ImageJ. At least 50 GFP/HA-positive nuclei were counted per chamber and were quantified for containing more than ten nuclear GFP-eta foci. Error bars represent SD of the mean, n = 3. MEFs MEFs from E13.5 embryos obtained from crosses between Rosa26 CRE-ER mice (Jackson ImmunoResearch Laboratories) and mice carrying a conditional TRF2 allele (Celli and de Lange, 2005) were immortalized at passage 2 with pBabeSV40LT. Ubiquitin-HA constructs were introduced by retroviral infection using the following constructs: pLPC-HA-Ub and pLPC-HAUb-L73P. IR treatments, metaphase spreads, and FISH/IF costaining were performed as described previously (Okamoto et al., 2013); 53BP1 was detected using a rabbit polyclonal antibody (Novus Biologicals, NB 100-304). Experiments were repeated in triplicates, and 53BP1 foci were

836 Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors

scored in at least 200 cells for each experiment. Data reported are averages of three independent experiments error bars represent SD. Stable Cell Lines HeLa cells expressing doxycycline-inducible FLAG-Ub-WT and FLAG-Ub-L73P were generated using Flippase (Flp) recombination target (FRT)/Flp-mediated recombination technology in HeLa-T-rex Flp-in cells, as described previously (Tighe et al., 2008). Cells were induced with 1 mg/ml doxycycline for the indicated times to induce transgene expression. For a medium scale mass spectrometry study, Flp-in UbL73P cells (36 3 15 cm dishes) were grown to 60% confluency and induced with 1 mg/ml doxycycline for 48 hr. Cells were briefly pelleted and frozen at 80 C until further processing. FLAG immune Precipitation of Flp-in UbL73P Cells Cell pellets were lysed in mRIPA buffer (20mM Tris [pH 7.5], 1% NP-40, 0.5% CHAPS, 0.1% SDS, and 150 mM NaCl) supplemented with protease inhibitor cocktail (Roche) and 1 U/ml benzoase (Novagen), for 1 hr on ice. Lysates were cleared by centrifugation and the supernatant was incubated 100 mg M2conjugated (anti-FLAG) magnetic DynaBeads (Sigma-Aldrich) for 1 hr at 4 C. The beads were extensively washed in mRIPA buffer, then in 100 mM ammonium bicarbonate (pH 8) and immediately processed for massspectrometry-grade trypsin digest. Mass Spectrometry Immunoprecipitated proteins were reduced with the addition of 0.02 M dithiothreitol (pH 8) for 1 hr at 57 C and subsequently alkylated using 0.05 M iodoacetic acid (pH 8) for 45 min in the dark at room temperature. The samples were proteolytically digested with trypsin (Promega) at a 1:20 enzyme to substrate ratio overnight with gentle shaking at room temperature. The resulting peptide mixture was removed from the beads and adjusted to pH 3 with trifluoroacetic acid prior to desalting using a C18 Stage tip procedure, as previously described (Cotto-Rios et al., 2012; Rappsilber et al., 2007; Wu et al., 2012). The desalted peptide mixture was concentrated in a SpeedVac concentrator to remove organic solvents. A tenth of the peptide mixture was loaded onto a Acclaim PepMap 100 precolumn (75 mm 3 2 cm, C18, 3 mm, 100 A˚, Thermo Scientific) that was connected to an EASY-Spray, PepMap RSLC column (75 mm 3 25 cm, C18, 2 mm, 100 A˚ Thermo Scientific) with a 5 mm emitter using the autosampler of an EASY-nLC 1000 (Thermo Scientific). Peptides were gradient eluted from the column directly into a Q Exactive mass spectrometer (Thermo Scientific) using a 120 min gradient from 2% solvent B to 40% solvent B. Solvent A was 2% acetonitrile in 0.5% acetic acid and solvent B was 90% acetonitrile in 0.5% acetic acid. High-resolution MS1 spectra were acquired with a resolution of 70,000, an AGC target of 1e6, with a maximum ion time of 120 ms, and scan range of 400 to 1,500 m/z. Following each MS1, 20 data-dependent high-resolution HCD MS2 spectra were acquired. All MS2 spectra were collected using the following instrument parameters: resolution of 17,500, AGC target of 5e4, maximum ion time of 250 ms, one microscan, 2 m/z isolation window, fixed first mass of 150 m/z, 30 s exclusion list, and NCE of 27. Database Search All MS2 spectra were searched against a uniprot human database using Sequest via Proteome Discoverer (Thermo Scientific). For the search trypsin was indicated as the enzyme, precursor mass tolerance was set to ±10 ppm, and fragment ion mass tolerance was set to ±0.02 Da. Carboxymethylation of Cys was searched as a static modification. The following variable modifications were also searched: diglycylation of Lys, deamidation of Gln and Asp, and oxidation of Met. Peptides identified as having a diglycylated lysine were manually validated.

SUPPLEMENTAL INFORMATION Supplemental Information includes four figures and one table and can be found with this article online at http://dx.doi.org/10.1016/j.celrep.2013.10.008.

AUTHOR CONTRIBUTIONS M.B., with input and help from T.T.H., conceived the project, designed and performed the experiments, analyzed the data, and wrote the manuscript. B.B.B. and F.D.M. synthesized the wild-type and (nonhydrolyzable, NH) Ub tetramers and contributed Figure 1G; K.O. and E.L.D. performed and analyzed the experiments with the TRF2F/F cells (Figures 4C–4E, S4B, and S4C); M.J.J. generated the Flp-in stable cells (used in Figure 5); S.B.C. cultured Flp-in cells and carried out the TNF-a experiment (Figure S4A) and aided in mass-spectrometry preparation. Protein identification by mass spectrometry was carried out by J.R.C. and B.M.U. ACKNOWLEDGMENTS The authors wish to thank Kristin Burns-Huang for critically reading the manuscript, members of the Lima, Sfeir, Smith, Reinberg, and Huang labs for reagents, resources, and discussions, and Shaun K. Olsen for the S. pombe E1 and for assistance with structure-based modeling. We express our gratitude to Chris Lima for hosting M.B. during the Hurricane Sandy lab relocation at NYU. M.B. is a recipient of an NRSA postdoctoral fellowship (1F32GM100598), and research in the Huang lab is supported by grants from the NIH (GM084244) and from ACS (RSG-12-158-DMC). E.L.D. is supported by the Pew Scholars Award. B.B.B. and F.D.M. are employees of Boston Biochem. Received: June 27, 2013 Revised: September 5, 2013 Accepted: October 3, 2013 Published: November 7, 2013 REFERENCES Be´ke´s, M., Prudden, J., Srikumar, T., Raught, B., Boddy, M.N., and Salvesen, G.S. (2011). The dynamics and mechanism of SUMO chain deconjugation by SUMO-specific proteases. J. Biol. Chem. 286, 10238–10247. Bohren, K.M., Gabbay, K.H., and Owerbach, D. (2007). Affinity chromatography of native SUMO proteins using His-tagged recombinant UBC9 bound to Co2+-charged talon resin. Protein Expr. Purif. 54, 289–294. Celli, G.B., and de Lange, T. (2005). DNA processing is not required for ATM-mediated telomere damage response after TRF2 deletion. Nat. Cell Biol. 7, 712–718. Cesare, A.J., Hayashi, M.T., Crabbe, L., and Karlseder, J. (2013). The telomere deprotection response is functionally distinct from the genomic DNA damage response. Mol. Cell 51, 141–155. Cotto-Rios, X.M., Bekes, M., Chapman, J., Ueberheide, B., and Huang, T.T. (2012). Deubiquitinases as a signaling target of oxidative stress. Cell Reports 2, 1475–1484. Dayal, S., Sparks, A., Jacob, J., Allende-Vega, N., Lane, D.P., and Saville, M.K. (2009). Suppression of the deubiquitinating enzyme USP5 causes the accumulation of unanchored polyubiquitin and the activation of p53. J. Biol. Chem. 284, 5030–5041. Denchi, E.L., and de Lange, T. (2007). Protection of telomeres through independent control of ATM and ATR by TRF2 and POT1. Nature 448, 1068–1071. Dou, H., Buetow, L., Sibbet, G.J., Cameron, K., and Huang, D.T. (2012). BIRC7-E2 ubiquitin conjugate structure reveals the mechanism of ubiquitin transfer by a RING dimer. Nat. Struct. Mol. Biol. 19, 876–883. Ernst, A., Avvakumov, G., Tong, J., Fan, Y., Zhao, Y., Alberts, P., Persaud, A., Walker, J.R., Neculai, A.M., Neculai, D., et al. (2013). A strategy for modulation of enzymes in the ubiquitin system. Science 339, 590–595. Giannakopoulos, N.V., Luo, J.K., Papov, V., Zou, W., Lenschow, D.J., Jacobs, B.S., Borden, E.C., Li, J., Virgin, H.W., and Zhang, D.E. (2005). Proteomic identification of proteins conjugated to ISG15 in mouse and human cells. Biochem. Biophys. Res. Commun. 336, 496–506. Hershko, A., and Ciechanover, A. (1998). The ubiquitin system. Annu. Rev. Biochem. 67, 425–479.

Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors 837

Hodgins, R.R., Ellison, K.S., and Ellison, M.J. (1992). Expression of a ubiquitin derivative that conjugates to protein irreversibly produces phenotypes consistent with a ubiquitin deficiency. J. Biol. Chem. 267, 8807–8812.

Rappsilber, J., Mann, M., and Ishihama, Y. (2007). Protocol for micro-purification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips. Nat. Protoc. 2, 1896–1906.

Hoege, C., Pfander, B., Moldovan, G.L., Pyrowolakis, G., and Jentsch, S. (2002). RAD6-dependent DNA repair is linked to modification of PCNA by ubiquitin and SUMO. Nature 419, 135–141.

Rawlings, N.D., Barrett, A.J., and Bateman, A. (2010). MEROPS: the peptidase database. Nucleic Acids Res. 38 (Database issue), D227–D233.

Hu, M., Li, P., Li, M., Li, W., Yao, T., Wu, J.W., Gu, W., Cohen, R.E., and Shi, Y. (2002). Crystal structure of a UBP-family deubiquitinating enzyme in isolation and in complex with ubiquitin aldehyde. Cell 111, 1041–1054.

Renatus, M., Parrado, S.G., D’Arcy, A., Eidhoff, U., Gerhartz, B., Hassiepen, U., Pierrat, B., Riedl, R., Vinzenz, D., Worpenberg, S., and Kroemer, M. (2006). Structural basis of ubiquitin recognition by the deubiquitinating protease USP2. Structure 14, 1293–1302.

Huang, T.T., Nijman, S.M., Mirchandani, K.D., Galardy, P.J., Cohn, M.A., Haas, W., Gygi, S.P., Ploegh, H.L., Bernards, R., and D’Andrea, A.D. (2006). Regulation of monoubiquitinated PCNA by DUB autocleavage. Nat. Cell Biol. 8, 339–347.

Reyes-Turcu, F.E., Ventii, K.H., and Wilkinson, K.D. (2009). Regulation and cellular roles of ubiquitin-specific deubiquitinating enzymes. Annu. Rev. Biochem. 78, 363–397.

Jackson, S.P., and Durocher, D. (2013). Regulation of DNA damage responses by ubiquitin and SUMO. Mol. Cell 49, 795–807. Jin, J., Li, X., Gygi, S.P., and Harper, J.W. (2007). Dual E1 activation systems for ubiquitin differentially regulate E2 enzyme charging. Nature 447, 1135–1138. Jones, M.J., Colnaghi, L., and Huang, T.T. (2012). Dysregulation of DNA polymerase k recruitment to replication forks results in genomic instability. EMBO J. 31, 908–918.

Sloper-Mould, K.E., Jemc, J.C., Pickart, C.M., and Hicke, L. (2001). Distinct functional surface regions on ubiquitin. J. Biol. Chem. 276, 30483–30489. Sobhian, B., Shao, G., Lilli, D.R., Culhane, A.C., Moreau, L.A., Xia, B., Livingston, D.M., and Greenberg, R.A. (2007). RAP80 targets BRCA1 to specific ubiquitin structures at DNA damage sites. Science 316, 1198–1202. Tighe, A., Staples, O., and Taylor, S. (2008). Mps1 kinase activity restrains anaphase during an unperturbed mitosis and targets Mad2 to kinetochores. J. Cell Biol. 181, 893–901.

Kannouche, P.L., Wing, J., and Lehmann, A.R. (2004). Interaction of human DNA polymerase eta with monoubiquitinated PCNA: a possible mechanism for the polymerase switch in response to DNA damage. Mol. Cell 14, 491–500.

Wang, B., Matsuoka, S., Ballif, B.A., Zhang, D., Smogorzewska, A., Gygi, S.P., and Elledge, S.J. (2007). Abraxas and RAP80 form a BRCA1 protein complex required for the DNA damage response. Science 316, 1194–1198.

Kim, H., Chen, J., and Yu, X. (2007). Ubiquitin-binding protein RAP80 mediates BRCA1-dependent DNA damage response. Science 316, 1202–1205.

Wilkinson, K.D., Tashayev, V.L., O’Connor, L.B., Larsen, C.N., Kasperek, E., and Pickart, C.M. (1995). Metabolism of the polyubiquitin degradation signal: structure, mechanism, and role of isopeptidase T. Biochemistry 34, 14535– 14546.

Komander, D., and Rape, M. (2012). The ubiquitin code. Annu. Rev. Biochem. 81, 203–229. Lehmann, A.R., Niimi, A., Ogi, T., Brown, S., Sabbioneda, S., Wing, J.F., Kannouche, P.L., and Green, C.M. (2007). Translesion synthesis: Y-family polymerases and the polymerase switch. DNA Repair (Amst.) 6, 891–899. Machida, Y.J., Machida, Y., Chen, Y., Gurtan, A.M., Kupfer, G.M., D’Andrea, A.D., and Dutta, A. (2006). UBE2T is the E2 in the Fanconi anemia pathway and undergoes negative autoregulation. Mol. Cell 23, 589–596. Okamoto, K., Bartocci, C., Ouzounov, I., Diedrich, J.K., Yates, J.R., 3rd, and Denchi, E.L. (2013). A two-step mechanism for TRF2-mediated chromosome-end protection. Nature 494, 502–505. Olsen, S.K., and Lima, C.D. (2013). Structure of a ubiquitin E1-E2 complex: insights to E1-E2 thioester transfer. Mol. Cell 49, 884–896. Petroski, M.D., and Deshaies, R.J. (2005). In vitro reconstitution of SCF substrate ubiquitination with purified proteins. Methods Enzymol. 398, 143–158. Plechanovova´, A., Jaffray, E.G., Tatham, M.H., Naismith, J.H., and Hay, R.T. (2012). Structure of a RING E3 ligase and ubiquitin-loaded E2 primed for catalysis. Nature 489, 115–120. Povlsen, L.K., Beli, P., Wagner, S.A., Poulsen, S.L., Sylvestersen, K.B., Poulsen, J.W., Nielsen, M.L., Bekker-Jensen, S., Mailand, N., and Choudhary, C. (2012). Systems-wide analysis of ubiquitylation dynamics reveals a key role for PAF15 ubiquitylation in DNA-damage bypass. Nat. Cell Biol. 14, 1089– 1098.

Wu, D., Chapman, J.R., Wang, L., Harris, T.E., Shabanowitz, J., Hunt, D.F., and Fu, Z. (2012). Intestinal cell kinase (ICK) promotes activation of mTOR complex 1 (mTORC1) through phosphorylation of Raptor Thr-908. J. Biol. Chem. 287, 12510–12519. Xia, Z.P., Sun, L., Chen, X., Pineda, G., Jiang, X., Adhikari, A., Zeng, W., and Chen, Z.J. (2009). Direct activation of protein kinases by unanchored polyubiquitin chains. Nature 461, 114–119. Zeng, W., Sun, L., Jiang, X., Chen, X., Hou, F., Adhikari, A., Xu, M., and Chen, Z.J. (2010). Reconstitution of the RIG-I pathway reveals a signaling role of unanchored polyubiquitin chains in innate immunity. Cell 141, 315–330. Zhang, Y., Zhou, L., Rouge, L., Phillips, A.H., Lam, C., Liu, P., Sandoval, W., Helgason, E., Murray, J.M., Wertz, I.E., and Corn, J.E. (2013). Conformational stabilization of ubiquitin yields potent and selective inhibitors of USP7. Nat. Chem. Biol. 9, 51–58. Zhao, B., Bhuripanyo, K., Schneider, J., Zhang, K., Schindelin, H., Boone, D., and Yin, J. (2012). Specificity of the E1-E2-E3 enzymatic cascade for ubiquitin C-terminal sequences identified by phage display. ACS Chem. Biol. 7, 2027– 2035. Zhou, W., Zhu, P., Wang, J., Pascual, G., Ohgi, K.A., Lozach, J., Glass, C.K., and Rosenfeld, M.G. (2008). Histone H2A monoubiquitination represses transcription by inhibiting RNA polymerase II transcriptional elongation. Mol. Cell 29, 69–80.

838 Cell Reports 5, 826–838, November 14, 2013 ª2013 The Authors