Archives of Biochemistry and Biophysics 503 (2010) 207–212
`
`Contents lists available at ScienceDirect
`
`Archives of Biochemistry and Biophysics
`
`j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / y a b b i
`
`C-terminal region of USP7/HAUSP is critical for deubiquitination activity and
`contains a second mdm2/p53 binding site
`Jianhong Ma a,⇑, John D. Martin b, Yu Xue c, Leng A. Lor b, Karen M. Kennedy-Wilson b, Robert H. Sinnamon b,
`
`Thau F. Ho b, Guofeng Zhang b, Benjamin Schwartz b, Peter J. Tummino d, Zhihong Lai e
`a Analytical Sciences, GlaxoSmithKline, UE1125, 709 Swedeland Road, King of Prussia, PA 19406, USA
`b Biological Reagent and Assay Development, 1250 South Collegeville Road, Collegeville, PA 19426, USA
`c Biopharm Research Unit, GlaxoSmithKline, 709 Swedeland Road, King of Prussia, PA 19406, USA
`d Emerging Science DPU, GlaxoSmithKline, 1250 South Collegeville Road, Collegeville, PA 19426, USA
`e Worldwide Clinical Trials, 1000 Continental Drive, King of Prussia, PA 19406, USA
`
`a r t i c l e
`
`i n f o
`
`a b s t r a c t
`
`Article history:
`Received 1 July 2010
`and in revised form 22 August 2010
`Available online 15 September 2010
`
`Keywords:
`USP7
`HAUSP
`mdm2
`p53
`Deubiquitination
`Second binding site
`
`Introduction
`
`USP7, also known as the hepes simplex virus associated ubiquitin-specific protease (HAUSP), deubiquiti-
`nates both mdm2 and p53, and plays an important role in regulating the level and activity of p53. Here,
`we report that deletion of the TRAF-like domain at the N-terminus of USP7, previously reported to con-
`tain the mdm2/p53 binding site, has no effect on USP7 mediated deubiquitination of Ubn-mdm2 and Ubn-
`p53. Amino acids 208–1102 were identified to be the minimal length of USP7 that retains proteolytic
`activity, similar to full length enzyme, towards not only a truncated model substrate Ub-AFC, but also
`Ubn-mdm2, Ubn-p53. In contrast, the catalytic domain of USP7 (amino acids 208–560) has 50–700 fold
`less proteolytic activity towards different substrates. Moreover, inhibition of the catalytic domain of
`USP7 by Ubal is also different from the full length or TRAF-like domain deleted proteins. Using glutathi-
`one pull-down methods, we demonstrate that the C-terminal domain of USP7 contains additional binding
`sites, a.a. 801–1050 and a.a. 880–1050 for mdm2 and p53, respectively. The additional USP7 binding site
`on mdm2 is mapped to be the C-terminal RING finger domain (a.a. 425–491). We propose that the C-ter-
`minal domain of USP7 is responsible for maintaining the active conformation for catalysis and inhibitor
`binding, and contains the prime side of the proteolytic active site.
`Ó 2010 Elsevier Inc. All rights reserved.
`
`The ubiquitin–proteasome pathway plays an important role in
`regulating many biological processes, including cell cycle, differen-
`tiation, immune response, DNA repair, and apoptosis [1,2]. Ubiqui-
`tination is a highly dynamic and reversible process, regulated by
`multiple ubiquitin-conjugating enzymes and deubiquitinating en-
`zymes. Formation of an isopeptide bond between the C-terminal
`carboxylate group of ubiquitin and the e-amino group of a Lys res-
`idue in a substrate protein proceeds through a three-step cascade
`mechanism involving the ubiquitin-activating enzyme (E1), a ubiq-
`uitin-conjugating enzyme (E2), and a ubiquitin-protein ligase (E3).
`Substrate proteins can form conjugates with a single ubiquitin or
`ubiquitin chains at a single or multiple Lys residues [1,2].
`Ubiquitination of proteins can be reversed by the deubiquitinat-
`ing enzymes (DUBs) [3–5]. DUBs have important functions, such as
`rescuing substrate proteins from proteasomal degradation, recy-
`cling Ub, and controlling protein trafficking. There are at least five
`
`⇑ Corresponding author. Fax: +1 610 270 7100.
`
`E-mail address: jane.2.ma@gsk.com (J. Ma).
`
`0003-9861/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
`doi:10.1016/j.abb.2010.08.020
`
`distinct subclasses of DUBs, of which the Ub-specific proteases
`(USPs)1 constitute the largest family with more than 50 family
`members [3–5]. USPs are cysteine proteases containing conserved
`regions in their amino acid sequence surrounding the Cys, His, and
`Asp/Asn residues that form the catalytic triad. Beyond the catalytic
`domain, USPs often encode N-terminal and C-terminal extensions
`that may be important for substrate recognition and subcellular
`localization. The substrate specificity and physiological functions
`of most USPs are not yet understood.
`One of the best characterized USP is USP7 or HAUSP (herpes
`simplex virus associated ubiquitin-specific protease). USP7 was
`shown to directly bind to and deubiquitinate the p53 tumor
`suppressor protein [6]. Overexpression of USP7 resulted in the
`stabilization of p53 in cells. However, when USP7 was knocked
`out either by siRNA [7] or by homologous recombination [8],
`
`1 Abbreviations used: Ub, ubiquitin; DUB, deubiquitinating enzyme; USP, ubiquitin-
`specific protease; HAUSP, hepes simplex virus associated ubiquitin-specific protease;
`Ub-AFC, ubiquitin 7-amino-4-trifluoromethylcoumarin; Ubal, ubiquitin aldehyde;
`TRAF, TNF receptor associated factors; SDS–PAGE, sodium dodecyl sulfate–polyacryl-
`amide gel electrophoresis; GST, glutathione-S-transferase; a.a., amino acid.
`
`Post-Grant Review Petition for US 9,840,491
`EXHIBIT 1019
`Page 1
`
`EXHIBIT 7
`DELANSORNE DECLARATION
`
`214LT:20700:449503:1:ALEXANDRIA
`
`

`

`208
`
`J. Ma et al. / Archives of Biochemistry and Biophysics 503 (2010) 207–212
`
`stabilization of p53 was observed rather than the expected effect of
`destabilization. This paradox can be explained by USP7’s preferen-
`tial deubiquitination of mdm2, an E3 ubiquitin ligase which ubiqui-
`tinates p53 and thereby mediates its degradation. Knockdown of
`USP7 results in the stabilization of p53, induction of p53 responsive
`genes, and growth inhibition or apoptosis in cancer cells with func-
`tional p53 [7,8]. Because of the important role USP7 plays in regulat-
`ing the p53 pathway, inhibition of USP7’s protease activity or its
`interaction with mdm2 could be an effective way to reactivate p53
`for the treatment of cancer.
`USP7 is composed of 1102 amino acids and several distinct do-
`mains. The N-terminal domain (NTD) of USP7 has sequence homol-
`ogy to the TNF receptor associated factors (TRAFs) and was shown
`to bind to mdm2, p53, as well as the Epstein–Barr virus nuclear
`antigen 1 (EBNA1) [9–13]. Crystal structures and binding studies
`suggest that the mdm2 peptides bind to the same surface groove
`in USP7 as peptides derived from p53, but with more extensive
`interactions and enhanced affinity [10,12]. The C-terminal region
`of USP7(560–1102) contains regions required for interactions with
`Ataxin-1 [14] and the herpes virus protein ICP0 [11]. The catalytic
`domain was mapped to amino acids 208–560 [9]. Crystal struc-
`tures of the catalytic domain in the presence and absence of a cova-
`lent inhibitor, ubiquitin aldehyde (Ubal), revealed that the active
`site of USP7 is misaligned in the apo structure. Formation of the
`covalent adduct of Ubal with the catalytic cysteine residue induces
`a drastic conformational change in the active site, realigning the
`catalytic triad for catalysis [9].
`In the current study, we have identified novel mdm2/p53 bind-
`ing sites on USP7 (a.a. 801–1050 and 880–1050 for mdm2 and p53,
`respectively), distinct from the N-terminal TRAF-like domain. We
`have also demonstrated that the catalytic domain of USP7 (a.a.
`208–560) is significantly less active than the full-length USP7 or
`truncated USP7 (a.a. 208–1102). Moreover, inhibition of the cata-
`lytic domain of USP7 is also different from the full length or
`TRAF-like domain deleted proteins. The C-terminus of USP7 ap-
`pears to be critical for its catalytic activity either by constituting
`part of the active site or by maintaining the active conformation
`for catalysis and for inhibitor binding.
`
`Material and methods
`
`Protein expression and purification of USP7, mdm2, and p53
`
`Full-Length USP7 (NP_003461) was cloned from a human colon
`cDNA library using standard PCR techniques. It was subcloned into
`the vector pENTR/TEV/D-TOPO (Invitrogen K252520) and used as a
`template to generate different USP7 truncations. All USP7 con-
`structs contained a 6 his tag and a tobacco etch virus (tev) prote-
`ase cleavage site N-terminal of the USP7 coding region. Protein
`expression was carried out in Sf9 insect cells with baculovirus
`infection at 27 °C for 3 days.
`Sf9 cells expressing various USP7 constructs were centrifuged
`and cell pellets were lysed in buffer A (20 mM Tris–HCl at pH
`7.6, 2 mM TCEP, 5% glycerol) plus 250 mM NaCl, 20 mM imidazole
`and EDTA free protease inhibitor cocktail (Roche). The cellular ex-
`tract was clarified by centrifugation at 17,000 rpm for 60 min at
`4 °C, before loading onto a Ni–NTA column. After washing exten-
`sively with buffer A plus 250 mM NaCl containing 20 mM and then
`40 mM imidazole, his-tagged USP7 was eluted with buffer A
`containing 250 mM imidazole and 250 mM NaCl. Subsequently,
`USP7 proteins were purified by size exclusion chromatography
`(superdex 200 in buffer A with 150 mM NaCl), followed by MonoQ
`chromatography (eluted in buffer A at 180 mM NaCl).
`GST-mdm2 and its deletion mutants, GST- or strep-tagged p53
`and GST-E1 were expressed and purified as described [15].
`
`Generation and partial purification of Ubn-p53 and Ubn-mdm2
`
`UbcH5b was conjugated to Cy5-labeled ubiquitin (Cy5-Ub-
`UbcH5b) as described [15]. Ubiquitinated p53 (Ubn-p53) was gen-
`erated using p53 and Cy5-Ub-UbcH5b in an enzymatic reaction
`catalyzed by mdm2. The reaction was carried out in buffer B
`(15 mM HEPES at pH 7.5, 5 mM NaCl, and 10 mM octylglucoside)
`containing 20 lM Cy5-Ub-UbcH5b, 5 lM strep-tagged p53, and
`0.2 lM GST-tagged mdm2 at room temperature for 60 min. To
`get rid of the excess Ub-UbcH5b and mdm2, strep-tagged p53 spe-
`cies were partially purified from the reaction mixture using stept-
`actin Sepharose resin (Genosys Biotech), washed with buffer C
`(50 mM Tris at pH 8.0, 150 mM NaCl, 10% glycerol, 0.1% triton
`X-100 amd 0.5 mM TCEP) with 1% triton X-100 followed by buffer
`C with 0.1% triton X-100, and eluted with 4 mM d-biotin. Partial
`purified Ubn-p53 was desalted using a G-25 column into buffer C
`without triton X-100. Mdm2-catalyzed p53 ubiquitination usually
`results in multiple monoubiquitinated p53 (Ubn-p53) containing
`mainly p53 with 1–5 ubiquitin conjugates [16]. Using a Cy5-Ub
`calibration curve, we analyzed the partially purified Ubn-p53 by
`SDS–PAGE, and calculated the concentration of Ubn-p53 based on
`the fluorescence intensity of Cy5-Ub that is attached to p53. We
`estimated that the partially purified Ubn-p53 mixture contained
`24% ubiquitinated p53.
`Ubn-mdm2 was generated using the mdm2 auto-ubiquitination
`reaction carried in buffer B containing 20 lM Cy5-Ub-UbcH5b and
`0.4 lM GST-tagged mdm2 for 4 h at room temperature. The reac-
`tion mixture was incubated with glutathione–Sepharose 4B resin
`(GE Healthcare), washed with buffer D (25 mM HEPES at pH 7.3,
`150 mM NaCl, 10% glycerol, and 0.1 mM TCEP) with 1% followed
`by 0.1% triton X-100, and finally eluted with buffer D containing
`0.1% triton X-100 and 20 mM glutathione. Partial purified Ubn-
`mdm2 was desalted into a buffer containing 20 mM HEPES at pH
`7.3, 1 mM DTT and 0.05% CHAPS using a G-25 column. Because
`mdm2 auto-ubiquitination usually results in a mixture of mdm2
`with high molecular weight ubiquitin chains, we quantified the
`amount of mdm2 before and after the reaction using western blot
`analysis with a monoclonal anti-mdm2 antibody (BD Pharmingen,
`#556353). Our estimate is that 70% of mdm2 has been converted
`to Ubn-mdm2 under our reaction conditions. In our USP7 mediated
`deubiquitination assays using Ubn-mdm2 as substrates, total pro-
`tein concentration of Ubn-mdm2 was indicated in the assays.
`
`In vitro ubiquitination assays for USP7
`
`We used three substrates to assess the activity of full-length
`USP7 and various truncated constructs: ubiquitin-7-amino-4-(tri-
`fluoromethyl)coumarin (ubiquitin-AFC, Boston Biochem), Ubn-p53
`and Ubn-mdm2. Ubiquitin-AFC cleavage was monitored fluoromet-
`rically, similar to procedures described for Ub-AMC [17]. Reactions
`were carried out in buffer E (50 mM HEPES pH 7.5, 0.5 mM EDTA,
`0.5 mM TCEP, 0.1 mg/ml BSA and 0.05% CHAPS) using low volume
`384 black assay plates (GreinerBio) at room temperature. The reac-
`tion mixture contains Ub-AFC and different USP7 constructs at var-
`ious concentrations. Cleavage of Ub-AFC by USP7 was monitored
`kinetically for 1 h using the analyst plate reader (Molecular Devices)
`for fluorescence intensity at excitation/emission wavelength of
`405 nm and 520 nm, respectively. The initial velocity of the reac-
`tions was used for steady-state analysis and IC50 determination.
`An SDS–PAGE assay was used to quantify USP7 mediated
`deubiquitination of Ubn-mdm2 and Ubn-p53. Reactions were car-
`ried out at room temperature using 2.65 lg/ml Ubn-p53 or
`1.68 lg/ml Ubn-mdm2, and USP7 enzymes at various concentra-
`tions in buffer E. After 20 min, the reactions were stopped with
`4 SDS-reducing sample buffer and separated using 4–12%
`NuPAGE Bis–Tris SDS–PAGE (Invitrogen). Fluorescence intensity
`
`Post-Grant Review Petition for US 9,840,491
`EXHIBIT 1019
`Page 2
`
`EXHIBIT 7
`DELANSORNE DECLARATION
`
`214LT:20700:449503:1:ALEXANDRIA
`
`

`

`J. Ma et al. / Archives of Biochemistry and Biophysics 503 (2010) 207–212
`
`209
`
`of Cy5-Ub product bands were quantified on a Typhoon 9400 im-
`ager (GE Healthcare) and converted to molar units using a Cy5-
`Ub calibration curve.
`A covalent USP7 inhibitor, ubiquitin aldehyde (Ubal) (Boston
`Biochem) [18], in serial dilution ranging from 25 lM to 0.005
`nM, was premixed with various USP7 constructs indicated above
`for 1 h at room temperature. Reactions were initiated with sub-
`strates (final concentration: 0.5 lM Ub-AFC; or 2.65 lg/ml Ubn-
`p53; or 1.68 lg/ml Ubn-mdm2) and carried out as described above.
`
`Binding interaction of USP7 to mdm2 or p53 by GST pull-down
`
`Various USP7 constructs (0.5 lM) were incubated with 0.5 lM of
`GST-p53, GST-mdm2, GST-E1 or GST in buffer F (50 mM HEPES pH
`7.5, 0.5 mM EDTA, 0.5 mM TCEP, 0.1 mg/ml BSA, 50 mM NaCl, and
`0.3% CHAPS) at room temperature for 30 min. GST-E1 and GST were
`included as negative controls for the pull-down. Glutathione–Se-
`pharose 4B resins (washed with buffer F, blocked with 0.5% nonfat
`milk and washed with buffer F again) were added and allowed to
`incubate for 1 h at 4 °C with mixing. The resins were washed four
`times with washing buffer (50 mM HEPES pH 7.5, 0.5 mM EDTA,
`0.5 mM TCEP, 0.1 mg/ml BSA, 150 mM NaCl, and 0.5% nonidet P-
`40) and heated for 10 min at 95 °C in 1 SDS–TrisGlycine sample
`loading buffer. Bound proteins were separated on 4–20% TrisGlycine
`polyacrylamide gels, transferred onto nitrocellulose membrane
`(Invitrogen), and analyzed by western blot using a monoclonal
`anti-his tag primary antibody (9.25 lg/ml) against his-tagged
`USP7 constructs and an Alexa Fluor 680 conjugated goat anti-mouse
`secondary antibody (0.8 lg/ml) (Molecular Probes); or a polyclonal
`anti-USP7 (1050–1102) primary antibody (1 lg/ml) (Bethyl Lab)
`and an Alexa Fluor 680 conjugated goat anti-rabbit secondary anti-
`body (0.8 lg/ml) (Molecular Probes). Immunodetection was carried
`out using the Odyssey Infrared Image system (Li-Cor Bioscience),
`and quantified by ImageQuant (Molecular Dynamics).
`
`Results
`
`Deletion mapping of USP7 for deubiquitination activity
`
`To understand the domain requirements for USP7’s deubiquiti-
`nation activity, we expressed and purified a series of USP7 con-
`structs as N-terminal His-tagged fusion proteins. All the truncated
`USP7 enzymes contain the catalytic domain (a.a. 208–560) with
`various C-terminal extensions,
`including 208–1102, 208–1050,
`208–880, 208–801 and 208–560. The N-terminal TRAF-like domain,
`previously reported to contain the mdm2 and p53 interaction sites,
`has been deleted from these truncated USP7 constructs.
`The deubiquitination activity of all of the truncated USP7 con-
`structs were compared with full-length USP7 using Ub-AFC as a
`model substrate (Table 1). USP7(208–1102) has similar activity
`as the full length protein. However, deletion of the last 52 amino
`acids results in 65-fold loss of activity based on kcat/Km for
`
`Table 1
`Kinetic constants of full-length USP7 and its deletion mutants for the hydrolysis of
`Ub-AFC.
`
`Enzyme
`
`USP7(1–1102)
`USP7(208–1102)
`USP7(208–1050)
`USP7(208–880)
`USP7(208–801)
`USP7(208–560)
`
`Ub-AFC
`kcat (s1)
`1.64 ± 0.07
`3.28 ± 0.14
`0.014 ± 0.001
`0.042 ± 0.005
`0.010 ± 0.0004
`0.014 ± 0.0005
`
`Km  106 (M)
`0.70 ± 0.03
`1.30 ± 0.09
`0.38 ± 0.12
`1.67 ± 0.11
`1.76 ± 0.19
`4.79 ± 0.70
`
`kcat/Km  106 (s1 M1)
`2.34 ± 0.007
`2.53 ± 0.067
`0.036 ± 0.008
`0.025 ± 0.001
`0.006 ± 0.004
`0.003 ± 0.0003
`
`Each kinetic constant is the mean of two replicates ± SD.
`
`USP7(208–1050). Deletion of additional amino acids from the
`C-terminus causes further decrease in activity. The loss of activity
`is primarily due to a decrease in kcat. While Km of Ub-AFC at 1 lM
`is very similar for various USP7 constructs (except USP7(208–560)
`which has a slightly higher Km (4.79 lM)), kcat is significantly dif-
`ferent and reflects the changes in the specific activity (kcat/Km)
`for the various USP7 constructs. These results suggest that deletion
`of C-terminal region of USP7 reduces catalysis more significantly
`than affecting the binding of Ub-AFC to the enzyme.
`Since Ub-AFC is a truncated model substrate for USP7, we gen-
`erated two physiological substrates (Ubn-mdm2 and Ubn-p53) to
`compare the specific activity of the various USP7 constructs. An
`SDS–PAGE assay was used to quantify USP7 mediated deubiquiti-
`nation of Ubn-mdm2 and Ubn-p53. Fluorescence intensity of
`Cy5-Ub product bands resulted from the deubiquitination of
`Ubn-mdm2 and Ubn-p53 were quantified on a Typhoon 9400 ima-
`ger (see Supplementary data). Our results showed that USP7(208–
`1102), lacking the N-terminal mdm2/p53 binding domain, has
`similar activity as full-length USP7 in catalyzing deubiquitination
`of not only Ub-AFC, but also Ubn-mdm2 and Ubn-p53 (Table 2).
`Similar trends of activity towards these substrates were also
`observed for the C-terminally deleted constructs. Deletion of the
`last 52 residues from the C-terminus results in significant
`(>20-fold) loss of activity and additional deletions further decrease
`the specific activity towards all three substrate tested. The catalytic
`core domain, USP7(208–560), is >700-fold less active than full-
`length USP7 towards Ub-AFC and Ubn-p53 substrates and 50-fold
`less active towards Ubn-mdm2.
`
`Inhibition of USP7 catalyzed deubiquitination by Ubal
`
`Since the deletion of the C-terminal region of USP7 resulted in a
`dramatic decrease of
`its proteolytic activity, we investigated
`whether inhibition of these USP7 constructs may also be different.
`Ubal is a potent inhibitor of USPs because it forms a covalent thio-
`hemiacetal adduct with the active site Cys [18]. We demonstrated
`that Ubal exhibits time-dependent inhibition of USP7 (data not
`shown). To take into consideration the time dependency of Ubal
`modification, we preincubated USP7 enzymes with Ubal for 1 h
`prior to substrate addition and measured IC50 rather than generat-
`ing kinact and Ki values. These IC50 values reflect an approximation
`of Ubal potency under our experimental conditions. Our results
`showed that Ubal is a very potent inhibitor against full-length
`USP7 and USP7(208–1102) with IC50 against Ub-AFC of 1.18 and
`1.50 nM, respectively (Table 3). However, the inhibitory effect of
`Ubal becomes weaker towards the C-terminally deleted USP7 con-
`structs. For example, using Ub-AFC as a substrate, IC50 of Ubal
`against USP7(208–560) core catalytic domain is 317 nM, 264-fold
`less potent than against full-length USP7. Similar decrease in Ubal
`potency against USP7(208–560) was also observed using the
`
`Table 2
`Comparison of relative activity of full-length USP7 and deletion mutants in catalyzing
`Ub-AFC, Ubn-p53 and Ubn-mdm2 deubiquitination.
`
`Enzyme
`
`USP7(1–1102)
`USP7(208–1102)
`USP7(208–1050)
`USP7(208–880)
`USP7(208–801)
`USP7(208–560)
`
`Substrates
`
`Ub-AFC
`
`100 ± 0.3
`108 ± 2.9
`1.56 ± 0.34
`1.07 ± 0.06
`0.25 ± 0.02
`0.12 ± 0.01
`
`Ubn-p53
`
`100 ± 11
`64 ± 8
`2.1 ± 0.3
`2.2 ± 0.2
`0.73 ± 0.09
`0.14 ± 0.02
`
`Ubn-mdm2
`
`100 ± 8
`118 ± 13
`4.4 ± 0.4
`11.3 ± 1.6
`0.66 ± 0.05
`2.0 ± 0.3
`
`Relative activity of full-length USP7 was set as 100. Specific activities of all USP7
`constructs were compared with the full-length USP7. Each relative activity is the
`mean of more than two replicates ± SD.
`
`Post-Grant Review Petition for US 9,840,491
`EXHIBIT 1019
`Page 3
`
`EXHIBIT 7
`DELANSORNE DECLARATION
`
`214LT:20700:449503:1:ALEXANDRIA
`
`

`

`210
`
`J. Ma et al. / Archives of Biochemistry and Biophysics 503 (2010) 207–212
`
`Table 3
`Effects of ubiquitin aldehyde on USP7 catalyzed hydrolysis of various substrates.
`
`Enzyme
`
`USP7(1–1102)
`USP7(208–1102)
`USP7(208–1050)
`USP7(208–880)
`USP7(208–801)
`USP7(208–560)
`
`Substrate
`
`Ub-AFC
`IC50 (nM)
`
`1.18 ± 0.12
`1.50 ± 0.10
`8.69 ± 1.00
`5.18 ± 1.21
`28.3 ± 5.1
`316.9 ± 6.4
`
`Ubn-p53
`IC50 (nM)
`
`1.02 ± 0.15
`
`–
`–
`–
`–
`433.8 ± 20
`
`Ubn-mdm2
`IC50 (nM)
`
`1.58 ± 0.61
`–
`–
`–
`–
`340.6 ± 224.2
`
`The data is the mean of two determinations ± SD.
`
`physiological substrates, Ubn-p53 and Ubn-mdm2. Taken together,
`these data demonstrate that inhibition by Ubal of USP7 catalyzed
`deubiquitination is less potent towards C-terminally truncated vs
`full-length USP7 protein.
`
`Mapping of the second mdm2/p53 binding site on USP7 critical for
`deubiquitination
`
`We have observed that USP7(208–1102), lacking the N-terminal
`mdm2/p53 binding TRAF-like domain, has similar activity as full-
`length USP7 in catalyzing deubiquitination of Ubn-mdm2 and
`Ubn-p53 (Table 2). This result suggested that binding through the
`N-terminal TRAF-like domain of USP7 is not essential for USP7 cat-
`alyzed mdm2 or p53 deubiquitination. Therefore, USP7 probably
`contain a second mdm2 and p53 binding site potentially critical
`for its deubiquitination activity.
`To probe for the putative interaction site between USP7
`constructs and mdm2 or p53, we carried out GST pull-down
`
`experiments. Our results showed that USP7 constructs 1–1102,
`208–1102 and 208–1050 bind to mdm2 and p53 (Fig. 1A and B).
`In addition, USP7(208–880) binds mdm2, but not p53. In contrast,
`USP7(208–801) and the catalytic domain USP7(208–560) does not
`bind to either mdm2 or p53. Therefore, 801–1050 residues of USP7
`contain the second binding site for mdm2; and 880–1050 residues
`of USP7 contain the second binding site for p53.
`
`Mapping of the USP7 binding site on mdm2 critical for
`deubiquitination
`
`To identify the additional site on mdm2 critical for USP7 bind-
`ing, we performed GST pull-down experiments using USP7 or
`USP7(208–1102) and various mdm2 deletion mutants. A series of
`truncated mdm2 constructs containing the C-terminal RING finger
`region with different deletion at the N-terminus were expressed
`and purified as N-terminal GST-tagged fusion proteins [15]. Our re-
`sults showed that both full-length USP7 and USP7(208–1102),
`lacking the previously identified N-terminal p53/mdm2 binding
`domain, can bind to full-length mdm2, mdm2-S211 (a.a. 211–
`491), mdm2-E285 (a.a. 285–491), mdm2-D361 (a.a. 361–491)
`and mdm2-S425 (a.a. 425–491) (Fig. 2); and as expected have no
`significant binding affinity for the negative control GST-E1 and
`GST proteins. Therefore, the RING finger domain (a.a. 425–491)
`contains the additional USP7 binding site.
`
`Discussion and conclusion
`
`USP7 is an important regulator of the mdm2/p53 pathway. It
`preferentially deubiquitinates mdm2 over p53. The net result of
`
`Binding to USP7
`by GST pulldown
`Activity p53 mdm2
`
`USP7
`
`1-205
`
`208-560
`
`560-880
`
`880-1102
`
` +++++ + +
`
`USP7(208-1102)
`
`208-560
`
`560-880
`
`880-1102
`
` +++++ +
`
` +
`
`USP7(208-1050)
`
`208-560
`
`560-880
`
`880-1102
`
` +++ + +
`
`USP7(208-880)
`
`208-560
`
`560-880
`
` ++ -
`
`USP7(208-801)
`
`208-560
`
`560-801
`
` + -
`
`USP7(208-560)
`
`208-560
`
` + -
`
` +
`
` -
`
` -
`
`A
`
`B
`
`USP7 1-1102 208-1102 208-1050
`1 2 3 4 1 2 3 4 1 2 3 4 MW
`kDa
`250
`
`USP7 208-880 208-801 208-560
`1 2 3 4 1 2 3 4 1 2 3 4 MW
`kDa
`250
`
`98
`64
`50
`36
`
`16
`6
`
`98
`64
`50
`36
`
`16
`6
`
`Fig. 1. Identification of the second p53 and mdm2 binding site on USP7. (A) A summary of USP7 activity and binding to p53 and mdm2 in various USP7 constructs. The
`regions of USP7 corresponding to the p53/mdm2 binding (1–205), the catalytic domain (205–560), the ICP0 binding domain (205–880) and the C-terminal domain (880–
`1102) have been labeled using different colors. (B) Binding of various USP7 constructs to GST tagged p53 or mdm2 using GST pull-down as described in materials and
`methods. GST-E1 was included as a negative control. Bound USP7 protein was detected by western blot using an anti-His antibody and the Odyssey Infrared Image system.
`Lane 1, GST-p53; 2, GST-mdm2; 3, GST-E1; 4, 2 pmol full-length USP7 or 1 pmol of USP7 deletion mutants as controls for antibody detection. MW, molecular mass markers.
`(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
`
`Post-Grant Review Petition for US 9,840,491
`EXHIBIT 1019
`Page 4
`
`EXHIBIT 7
`DELANSORNE DECLARATION
`
`214LT:20700:449503:1:ALEXANDRIA
`
`

`

`J. Ma et al. / Archives of Biochemistry and Biophysics 503 (2010) 207–212
`
`211
`
`A
`
`GST p53 binding
`
`acidic domain Zn finger RING finger
`
`1 102 223 274 305 322 438 478 491
`
`Mdm2-FL
`
`Mdm2-S211
`
`Mdm2-E285
`
`Mdm2-D361
`
`Mdm2-S425
`
`B
`
`Full-length USP7
`
`USP7 (208-1102)
`
`1 2 3 4 5 6 7 8 MW
`
`1 2 3 4 5 6 7 8 MW
`
`Fig. 2. Identification of the second USP7 binding site on mdm2. (A) A summary of various GST-mdm2 constructs. The regions of mdm2 corresponding to the p53 binding, the
`acidic, the zinc finger and the RING finger domain have been labeled using different colors. The numbers corresponds to amino acid residues of mdm2. (B) Western blots of
`GST pull-down using full-length USP7 or USP7(208–1102) and GST–mdm2 as described in materials and methods. GST and GST–E1 were included as negative controls. Lane
`1, GST-mdm2; 2, GST-mdm2-S211; 3, GST-mdm2-E285; 4, GST-mdm2-D361; 5, GST-mdm2-S425; 6, GST; 7. GST-E1; 8, 1.6 pmol full-length USP7 or 1 pmol of USP(208–
`1102) mutant as controls for antibody detection. MW, molecular mass markers. (For interpretation of the references to colour in this figure legend, the reader is referred to the
`web version of this article.)
`
`USP7 knockout is the destabilization of mdm2 and induction of
`p53 and p53 responsive genes [7,8]. Targeting the USP7 deubiqui-
`tinating activity or the mdm2 interaction with USP7 could be effec-
`tive ways to treat cancer by activating wild type p53. Therefore, it
`is important to gain a better understanding of the mechanism of
`substrate recognition and catalysis for USP7. In this report, we have
`identified a novel mdm2/p53 binding site on USP7 (a.a. 801–1050/
`880–1050, respectively). We have also demonstrated that the cat-
`alytic domain of USP7(208–560) is significantly less active than the
`full-length USP7 or USP7(208–1102). Moreover, inhibition of the
`catalytic domain of USP7 by Ubal is also different from the full
`length or TRAF-like domain deleted proteins. The C-terminus of
`USP7 appears to be critical for its catalytic activity either by consti-
`tuting part of the active site or by maintaining the proper confor-
`mation for catalysis and inhibitor binding.
`The region of USP7 critical for mdm2 and p53 binding was
`originally mapped to the N-terminal TRAF-like domain by Hu
`and colleagues [9]. Using a large p53 fragment (a.a. 94–393),
`USP7(53–208) was identified to be the p53 binding motif by gel
`filtration analysis. They also identified a p53 fragment (a.a. 357–
`382) to be both necessary and sufficient for USP7(53–208) binding
`[9]. Using a large mdm2 fragment (a.a. 170–423), they reported
`that USP7(53–208) also contains a mdm2 binding site, and that a
`mdm2 fragment (a.a. 223–232) is essential to bind to USP7(53–
`208) [10]. The crystal structures of the USP7 TRAF-like domain
`alone and with a p53 peptide (359–368) or a mdm2 peptide
`(223–332) fused to its C-terminus showed that the p53 and
`mdm2 peptides recognize the same surface groove in USP7. How-
`ever, the mdm2 peptide has more extensive interactions and better
`affinity than the p53 peptide [10]. Sheng et al. also co-crystalized
`the TRAF-like domain of USP7 with peptides derived from p53
`(a.a. 359–367) and mdm2 (a.a. 141–150) [13]. They found that
`these peptides bind the same surface on USP7 as Epstein–Barr nu-
`clear antigen-1, explaining the competitive nature of the interac-
`tions. It is worth noting that while the two groups identified the
`same p53 peptide critical
`for interacting with the TRAF-like
`domain of USP7, the mdm2 peptides that each group has identified
`
`are different, suggesting that there are at least two binding sites on
`mdm2 for the TRAF-like domain of USP7.
`These earlier studies highlighted the importance of USP7’s
`N-terminal TRAF-like domain for mdm2/p53 interaction. Surpris-
`ingly, we demonstrated that the deletion of the TRAF-like domain
`had no impact on USP7’s deubiquitination activity towards not
`only Ub-AFC, but also Ubn-mdm2 and Ubn-p53. This paradox has
`led us to carry out GST pull-down experiments using full-length
`mdm2 and p53 to map their binding domains on USP7. We demon-
`strated that the C-terminal domain of USP7 contains a second site
`required for mdm2/p53 binding (a.a. 801–1050 and a.a. 880–1050
`for mdm2 and p53 binding, respectively). We speculate that the
`earlier studies have missed this second mdm2/p53 binding site
`on USP7 because fragments of p53 (missing the N-terminal do-
`main) and mdm2 (without the N-terminal p53 binding domain
`and the C-terminal RING finger domain) rather than full length
`proteins were used for complex formation with various USP7 con-
`structs. In fact, our deletion mapping of mdm2 identified the RING
`finger domain to contain the additional USP7 interaction site.
`Although we did not map the second USP7 binding site on p53, it
`is possible that p53 (1–94) contains the second USP7 binding site
`that interacts with USP7(880–1050).
`USP7, mdm2 and p53 proteins all contain at least two binding
`sites for each other [15,19]. Different binding sites correspond to
`different functions. The N-terminal interaction of mdm2 and p53
`results in the inhibition of p53 transcriptional activation, while
`the second binding between mdm2(211–361) and p53 core do-
`main is critical for mdm2-catalyzed p53 ubiquitination [15,19].
`For USP7, even though the N-terminal TRAF-like domain does bind
`to mdm2 and p53, it is not essential for USP7’s deubiquitinating
`activity towards mdm2 and p53. Instead, the TRAF-like domain
`has been reported to affect nuclear localization of USP7, suggesting
`that it may have functions not directly related to its enzymatic
`activity. The C-terminal region (a.a. 560–1102) has been reported
`to contain the binding site for protein partners like ICP0 and Atax-
`in-1 [11,14]. A possible role of the C-terminal region for oligomer-
`ization has also been suggested [20]. Our results demonstrated that
`
`Post-Grant Review Petition for US 9,840,491
`EXHIBIT 1019
`Page 5
`
`EXHIBIT 7
`DELANSORNE DECLARATION
`
`214LT:20700:449503:1:ALEXANDRIA
`
`

`

`212
`
`J. Ma et al. / Archives of Biochemistry and Biophysics 503 (2010) 207–212
`
`C-terminal region of USP7 contains a second p53/mdm2 binding
`site and is critical for catalysis, either by constituting part of the ac-
`tive site or by maintaining the active conformation for catalysis
`and for inhibitor binding. Since USP7’s deubiquitinating activity
`is extremely sensitive to the loss of the last 52 amino acids of
`the C-terminus beyond the second mdm2/p53 binding site, activity
`measurement of our deletion mutants did not allow us to conclude
`that the second mdm2/p53 binding site is indeed critical for sub-
`strate recognition and deubiquitination. Additional investigations
`will be required to unambiguously demonstrate whether the sec-
`ond mdm2/p53 binding sites critically contribute to substrate rec-
`ognition and catalysis.
`USP7(208–1102) is the minimal domain required for activity
`and substrate recognition. Within this region of USP7, a.a. 208–
`560 contains the catalytic triad and the ubiquitin binding site.
`USP7(801–1050) contains the p53/mdm2 binding site. Deletion
`of the last 52 residues of USP7 C-terminus has very dramatic im-
`pact on USP7 activity. Therefore, USP7(1050–1102) either forms
`part of the active site or maintains the active conformation of
`USP7. Inter