Background
Ubiquitin Specific Protease 1 (USP1) is a human deubiquitinase (DUB) that plays an important role in the regulation of the cellular response to DNA damage and is also involved in the control of cell differentiation (reviewed in [
1]). USP1 is a 785 amino acid protein, whose three-dimensional structure has not yet been solved. Structural analysis of other USP family members, such as USP7, has shown that the catalytic domain of these enzymes adopts a fold that resembles an open right hand with three sub-domains termed Fingers, Palm and Thumb [
2,
3]. A detailed sequence alignment analysis has further revealed that the USP core catalytic domain can be divided into six conserved boxes (boxes 1–6), and that several of these enzymes, including USP1, contain additional non-conserved domains inserted between the boxes that may play a regulatory role [
4]. USP1 bears one of the largest catalytic domains within the USP family, which includes two inserted domains between boxes 2 and 3, and between boxes 5 and 6 [
4].
One of the best-characterized functions of USP1 in the DNA damage response is as a regulator of Proliferating Cell Nuclear Antigen (PCNA) ubiquitination [
5]. Following DNA damage that stalls progression of the replication fork, PCNA is monoubiquitinated to promote the recruitment of translesion synthesis (TLS) DNA polymerases, which can bypass the lesion [
6]. USP1 deubiquitinates PCNA, thus contributing to prevent unscheduled recruitment of error-prone TLS DNA polymerases [
5].
USP1 carries out its function in the context of a heterodimeric complex with its cofactor USP1-Associated Factor 1 (UAF1). UAF1 has been shown to stabilize USP1 [
7] and to allosterically increase its catalytic activity, which is very low in the absence of the cofactor [
7,
8]. UAF1 also contributes to target USP1 to its nuclear substrates [
9]. In addition to USP1, UAF1 also binds to and regulates the activity of two other members of the USP family, USP12 and USP46 [
10].
Overexpression of USP1 has been reported in osteosarcoma [
11] and non-small cell lung cancer (NSCLC) [
1,
12], among other cancer types. In addition, USP1 mutations have been identified, albeit at a low frequency, in tumor samples [
13]. The functional effect of these cancer-associated USP1 mutations remains to be investigated. Importantly, several inhibitors of the USP1/UAF1 complex have been recently shown to act synergistically with cisplatin in cancer-derived cell lines [
1,
14,
15], suggesting that this complex may represent a valid therapeutic target in cancer. The development and implementation of USP1-targeted therapies will benefit from a more detailed knowledge of how the function of this DUB is regulated, and how this regulation can be affected by cancer-related mutations.
Several regulatory mechanisms converge to determine the levels, localization and activity of USP1 (reviewed in [
1]). These mechanisms include phosphorylation and autocleavage.
Phosphorylation at serine 313 (S313), within the first inserted domain, was initially reported to regulate cell cycle-dependent degradation of USP1. Thus, cyclin-dependent kinase 1 (CDK1)-mediated phosphorylation of S313 during M phase was shown to prolong the stability of USP1 presumably by preventing its degradation by the anaphase-promoting complex/cyclosome [
16]. More recently, S313 phosphorylation has been reported to be critical for UAF1 interaction and USP1 catalytic activity
in vitro [
17]. In this study, using pull-down assays with purified recombinant proteins, the UAF1-binding motif was mapped to USP1 amino acid region 235–408 [
17]. In contrast, using cell-based assays, we have mapped the UAF1-binding region in USP1 to a 100 amino acid motif comprising residues 420–520, which does not include the S313 phosphorylation site [
18]. It is, therefore, necessary to clarify these conflicting results, and to determine to what extent these two amino acid motifs, and the S313 phosphorylation site, contribute to USP1/UAF1 interaction.
Another important regulatory mechanism involves the autocleavage of USP1 at a diglycine motif (G670/G671) located within its second inserted domain [
5]. The cleavage event generates a longer N-terminal fragment (residues 1–671) and a shorter C-terminal fragment (residues 672–785), which are subsequently degraded by the proteasome [
5,
19]. It remains to be elucidated if USP1 autocleavage occurs in
cis (intramolecularly) or in
trans (intermolecularly) and, more importantly, if USP1 autocleavage could be altered by cancer-associated USP1 mutations that cluster around the G670/G671 motif.
In the present work, we use site-directed mutagenesis and cell-based functional assays to carry out a detailed structure-function analysis of human USP1. Our results indicate that the S313 phosphorylation site is not critical for USP1 binding to UAF1 or for PCNA deubiquitination in a cellular environment. Furthermore, we show that two homologous amino acid segments in USP1 (420–520) and USP46 (165–259), which are predicted to correspond to the Fingers sub-domains, mediate binding of these DUBs to UAF1. Moreover, we provide some experimental evidence suggesting that USP1 autocleavage may occur in cis. Finally, we identify a cancer-associated mutation in a residue adjacent to the cleavage site (L669P), that hampers USP1 autocleavage.
Discussion
Since the discovery, nearly a decade ago, that USP1 plays an important role in the cellular response to DNA damage [
5,
25], significant advances have been made on the understanding of the function and regulation of this DUB. It has been shown, for example, that USP1 carries out its cellular activities in the context of a heterodimeric complex with UAF1, which enhances the stability and the enzymatic activity of the DUB [
7,
8]. The stability and activity of USP1 have been shown to be further regulated by several mechanisms, including CDK-mediated phosphorylation at the S313 residue [
16] and cleavage by either itself [
5] or other proteases [
23,
26].
The USP1/UAF1 complex is emerging as a novel target for cancer treatment (reviewed in [
1]), and inhibitors of USP1 catalytic activity have been reported to reverse the resistance to platinum-based chemotherapeutic drugs in NSCLC cells [
14,
15] and to inhibit the growth of leukemic cell lines [
27]. Further development of the therapeutic potential of this deubiquitinase complex would benefit from a better understanding of several aspects of its regulation that remain controversial, incompletely characterized or unexplored.
For example, protein-protein interactions are increasingly regarded as promising therapeutic targets in cancer [
28] and thus, characterizing the molecular determinants of USP1/UAF1 interaction may be important to guide efforts aimed to disrupt this interaction with therapeutic purposes. In this regard, phosphorylation at the S313 residue has been recently reported to be critical for USP1 activity and interaction with UAF1 in an
in vitro setting [
17]. Here, we show that a non-phosphorylatable mutant USP1
S313A is still able to bind UAF1 when tested in cell-based assays. These assays were carried out using ectopically expressed proteins, but are, in our view, closer to a physiological setting than
in vitro binding assays with purified proteins. In particular, the nuclear relocation assay with fluorescently-tagged proteins allows for visualization of the results in live cells. Of note, USP1/UAF1 interactions demonstrated using this assay could be consistently validated by co-IP in the present study, as well as in our previous report [
18]. Importantly, our data indicate that USP1
S313A forms a functional complex with UAF1 in transfected cells. This complex is able to reverse monoubiquitination of endogenous PCNA, a well-established cellular substrate of USP1 [
5,
15,
26,
29]. Our results do not definitely exclude the possibility that S313 phosphorylation may modulate the activity of the USP1/UAF1 complex. In fact, minor, but reproducible differences in the levels of ubPCNA were noted between cells expressing wild type USP1, USP1
S313A or a phosphomimetic USP1
S313D mutant. In line with previous
in vitro results using an artificial fluorogenic substrate [
17], PCNA deubiquitination appeared to be more efficient in cells expressing the S313D phosphomimetic USP1. Nevertheless, S313 phosphorylation does not seem to be an essential requisite for USP1 activity or UAF1 binding in a cellular environment.
Our finding that phosphorylation of USP1 at S313 is dispensable for UAF1 binding in cells, led us to use the relocation assay to directly compare the relative contribution of two proposed UAF1-binding sites in USP1: the 235–408 fragment containing the S313 residue [
17], and the 420–520 fragment [
18]. In marked contrast to the 420–520 fragment, the 235–408 fragment (in either the S313 wild type or phosphomimetic forms) was unable to promote the nuclear relocation of UAF1-mRFP. Furthermore, deletion of the 420–520 fragment abrogated UAF1 relocation, regardless of the presence of the S313D phosphomimetic mutation. Thus, USP1 420–520 amino acid motif is both necessary and sufficient for UAF1 binding in cells. It is important to note that this motif was not tested in the previous
in vitro UAF1-binding assays [
17].
The present findings seem to contradict previous
in vitro results showing an interaction of the USP1(235–408)
S313D motif with UAF1. It must be taken into account, however, that many interacting partners of both USP1 and UAF1 may be present within the crowded environment of intact cells [
30]. If USP1(235–408)
S313D interacts weakly with UAF1, a negative result in the nuclear relocation assay may be due to competition by endogenous partner(s). In this regard, nuclear import receptors (importins) were obvious candidates to compete with UAF1 for binding to USP1(235–408), because USP1 NLSs lie within this motif, and the importin KPNA1 has been identified as a potential USP1 interactor [
30]. Although USP1 also bears a potential nuclear export sequence (NES) that could mediate interaction with the export receptor CRM1, the physiological relevance of this sequence remains to be established [
13], and it lies outside the 235–408 segment. Importantly, NLS mutations do not increase nuclear relocation of UAF1-mRFP by USP1(235–408)
S313D, suggesting that competition with importins does not account for the lack of interaction between USP1(235–408)
S313D and UAF1 in a cellular setting.
Altogether, our findings indicate that the USP1 420–520 fragment is the critical site for robust UAF1-binding in a cellular environment. In line with this view, the UAF1-binding site of USP46, another UAF1-interacting DUB [
10], was mapped to amino acids 165–259, which is a motif homologous to USP1(420–520). The three-dimensional structure of USP1 and USP46 have not yet been solved, but
in silico modeling using the structure of USP7 catalytic domain [
2] as a template indicates that USP1(420–520) and USP46(165–259) UAF1-binding motifs lie within their Fingers sub-domain. Structural analyses have shown that the Fingers sub-domain contributes to ubiquitin binding by other members of the USP family [
2,
31,
32]. Since UAF1 stimulates the deubiquitinating activity of USP1 and USP46 [
7,
10], we speculate that UAF1 and ubiquitin can simultaneously bind to opposite surfaces of the Fingers sub-domain in these DUBs.
Besides S313 phosphorylation, a well-established mechanism regulating USP1 is autocleavage at the G670/G671 diglycine motif [
5]. Although originally reported to be induced by UV light [
5], we could readily detect this cleavage event in 293T cells co-expressing GFP-USP1 and Xpress-UAF1. This observation provided a convenient experimental system to evaluate two aspects of USP1 autocleavage that have not been tested. On one hand, we have used co-expression of catalytically inactive (USP1
C90S) and non-cleavable (USP1
GG/AA) mutants to provide some experimental evidence suggesting that USP1 may undergo autocleavage in
cis. Due to the technical limitations of our assay, this evidence is not conclusive. A conceptually similar
in vitro test, co-incubating purified recombinant forms of USP1
C90S and USP1
GG/AA, could provide further evidence. On the other hand, we have evaluated how USP1 autocleavage could be altered by mutations identified in human tumors.
To the best of our knowledge, no functional test on naturally-occurring USP1 mutations has yet been reported and thus, the effect that cancer-associated mutations may have on the function or regulation of USP1 remains unexplored. Around forty cancer-associated USP1 mutations were included in the COSMIC database by September 2013 [
13]. Some of these changes were non-sense or frameshift mutations that would most likely result in a non-functional allele, but most USP1 mutations lead to single amino acid substitutions whose effect is still unknown. Four of these missense mutations, G667A, L669P, K673T and A676T, are located adjacent or in close proximity to USP1 autocleavage site, and our data indicate that one of these changes, the L669P mutation, decreases USP1 cleavage efficiency. This region of USP1 corresponds to a non-conserved inserted domain between boxes 5 and 6, and therefore, no structural information can be obtained by
in silico modeling using other USPs as template. We hypothesize that this mutation might introduce a conformational change that hampers access of the cleavage site to the catalytic site. The USP1
L669P mutant clearly retains its ability to deubiquitinate PCNA in HU-treated cells. It is tempting to speculate that, by disrupting the normal balance of USP1 cleavage, this L699P mutation may contribute to tumorigenesis, and further experiments should address this possibility.
Methods
Plasmids, cloning procedures and site-directed mutagenesis
Plasmids encoding GFP-USP1 and Xpress-UAF1 were generously provided by Dr. Rene Bernards (Netherlands Cancer Institute, Amsterdam, The Netherlands) and Dr. Jae U. Jung (University of Southern California, Los Angeles, USA), respectively. Flag-HA-USP46 was obtained from the laboratory of Dr. John W. Harper (Harvard Medical School, Boston, USA) through Addgene (Plasmid #22584). YFP-USP1 (del420–520) and NLS-USP1(420–520)-GFP plasmids have been described previously [
18].
To generate the plasmid encoding UAF1-mRFP, UAF1 cDNA was amplified by PCR and cloned in frame to mRFP using XhoI/AgeI restriction sites. On the other hand, a DNA sequence encoding USP1 amino acid segment 235–408 was amplified by PCR using GFP-USP1 as a template, and cloned as either a KpnI/BamHI fragment into pEYFP-C1 (Clontech), or as a BamHI/AgeI fragment into a vector termed pNLS(SV40)-GFP. The vector pNLS(SV40)-GFP derives from a NES-GFP plasmid previously used in a nuclear import assay [
18]. Finally, DNA sequences encoding full length USP46 and the three deletion mutants (1–164, 165–259 and 243–366) were amplified by PCR using Flag-HA-USP46 as a template, and cloned as BamHI/AgeI fragments into pNLS(SV40)-GFP. Full-length USP46 was also cloned as a BamHI/AgeI fragment into a mutant version of pNLS(SV40)-GFP vector carrying a non-functional NLS sequence to generate the USP46-GFP plasmid. All PCR amplifications were carried out using high fidelity Pfu UltraII fusion HS DNA polymerase (Stratagene).
USP1 point mutations were created using the QuickChange Lightning Site-Directed Mutagenesis Kit (Stratagene), according to manufacturer’s directions.
All the new constructs generated were subjected to DNA sequencing (STABVIDA), and the absence of any unwanted mutation was confirmed. The sequences of the oligonucleotides used in cloning and site-directed mutagenesis are available upon request.
Cell culture, transfection and drug treatment
Human embryonic kidney 293T (HEK293T) cells were grown in Dulbecco’s modified Eagle’s medium (Invitrogen), supplemented with 10% fetal bovine serum (Invitrogen), 100 U/ml penicillin and 100 μg/ml streptomycin (Invitrogen). Twenty four hours before transfection cells were seeded in 12-well or 6-well tissue culture plates or 10 cm petri dishes. Transfections were carried out with X-tremeGENE 9 transfection reagent (Roche Diagnostics) following manufacturer’s protocol.
Hydroxyurea (Sigma-Aldrich) was added to the culture medium 48 hours after transfection to a final concentration of 4 mM for 24 hours.
Microscopy analysis of nuclear relocation assay samples
The nuclear relocation assay was carried out in cells seeded onto sterile glass coverslips. Cells co-expressing UAF1-mRFP with the different GFP- or YFP-tagged proteins were fixed with 3.7% formaldehyde in phosphate-buffered saline (PBS) for 30 min, incubated with Hoechst 33285 (Sigma) to visualize the nuclei, washed with PBS, and mounted onto microscope slides using Vectashield (Vector laboratories). Single-slice images were acquired using an Olympus Fluoview FV500 confocal microscope. Sequential acquisition of each fluorochrome was performed in order to avoid overlapping of fluorescent emission spectra. For for live cell imaging, cells were grown in 35 mm ibiTreat μ-dish slides (Ibidi), transfected with the indicated plasmids and examined using a Zeiss ApoTome.2 microscope. Semiquantitative analysis of nuclear relocation assay samples was carried out by determining the nucleocytoplasmic localization of UAF1-mRFP in at least 100 co-transfected cells per slide using a Zeiss Axioskop fluorescence microscope. Slides were coded to ensure unbiased scoring, and examined by two independent observers.
Immunoblot analysis and co-immunoprecipitation
Cells were washed with ice-cold PBS and collected in lysis buffer containing 10 mM sodium phosphate (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 10 mM β-glycerophosphate, 0.5% NP40, 10 mM phenylmethylsulfonyl fluoride, 10 mM sodium orthovanadate, 10 μg/ml protease inhibitor cocktail (Roche), and 50 mM N-ethylmaleimide (Thermo Scientific). Protein concentration was determined using the DC Protein Assay (Bio-Rad). For immunoblot analysis, protein samples were resolved in 8%, 10% or 12% SDS-PAGE and transferred onto nitrocellulose membranes (Bio-Rad). Prior to antibody incubation, membranes were stained with Ponceau to assess protein loading. Membranes were blocked with 5% non-fat dry milk in TTBS and incubated with the primary antibodies: anti-GFP (Chromotek, 1:1000), anti-Xpress (Invitrogen, 1:5000), anti-PCNA (Santa Cruz, 1:400), anti-β-actin (Sigma-Aldrich, 1:3000) and anti-α-tubulin (Sigma-Aldrich, 1:3000). Subsequently, membranes were incubated with the corresponding horseradish peroxidase-conjugated secondary antibodies (Santa cruz, 1:3000), and developed with ECL or Femto chemiluminiscence reagents (Thermo Scientific). Semiquantive analysis of immunoblot bands was performed by densitometry using Quantity One software 4.6 (Bio-Rad Laboratories).
Immunoprecipitation of GFP- or YFP-fusion proteins was carried out using Magnetic GFP-Trap beads (Chromotek), following manufacturer’s directions. Immunoprecipitated proteins were analysed by immunoblot as described above.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
AO-H designed and carried out experiments, and contributed to writing the manuscript. IG-S designed and carried out experiments, and contributed to writing the manuscript. JAR conceived the study, designed experiments and contributed to writing the manuscript. All authors read and approved the final manuscript.