DDB2 (damage-specific DNA-binding protein 2) is the product of the xeroderma pigmentosum group E gene which is involved in the initiation of nucleotide excision repair via an ubiquitin ligase complex together with DDB1 and CUL4A (cullin 4A). PAQR3 (progestin and adipoQ receptor family member III) is a newly discovered tumour suppressor that is implicated in the development of many types of human cancers. In the present paper, we report that DDB2 is involved in ubiquitination and degradation of PAQR3. DDB2 is able to interact with PAQR3 in vivo and in vitro. Both overexpression and knockdown experiments reveal that the protein expression level, protein stability and polyubiquitination of PAQR3 are changed by DDB2. Negative regulation of EGF (epidermal growth factor)- and insulin-induced signalling by PAQR3 is also altered by DDB2. At the molecular level, Lys61 of PAQR3 is targeted by DDB2 for ubiquitination. The cell proliferation rate and migration of gastric cancer cells are inhibited by DDB2 knockdown and such effects are abrogated by PAQR3 knockdown, indicating that the effect of DDB2 on the cancer cells is mediated by PAQR3. Collectively, our studies not only pinpoint that DDB2 is a post-translational regulator of PAQR3, but also indicate that DDB2 may play an active role in tumorigenesis via regulating PAQR3.
DDB2 (damage-specific DNA-binding protein 2) is the product of the XPE (xeroderma pigmentosum group E) gene and a defect in this gene causes an autosomal recessive disorder characterized by photosensitivity and early onset of carcinomas [1,2]. DDB2 forms a heterodimeric complex with DDB1 (damage-specific DNA-binding protein 1). DDB1 and DDB2 then form a complex with CUL4A (cullin 4A) and participate in the recognition of UV-damaged DNA and initiation of NER (nucleotide excision repair) [3–6]. In the DDB1–DDB2–CUL4A complex, DDB1 functions as an adaptor molecule and CUL4A is an ubiquitin ligase component. DDB2 serves as a substrate receptor module that determines the specificity of targeted substrate . During NER, DDB2 participates in the recognition of UV-induced pyrimidine photodimer . Upon the binding of the DDB1–DDB2–CUL4A complex to the UV-induced DNA damage site, the complex promotes ubiquitination of histones H2A, H3 and H4 to expose the damage sites [7,8]. Then the initiator of NER, XPC (xeroderma pigmentosum group C), is recruited to the damage sites to activate the NER process, while polyubiquitination of XPC by the complex alters the DNA-binding properties of XPC and is crucial for execution of NER [4,9].
In addition to its function in NER, DDB2 has been reported to participate in other biological processes, including carcinogenesis. DDB2 could associate with the histone acetyltransferases CBP [CREB (cAMP-response-element-binding protein)-binding protein]/p300 . Additionally, DDB2 has been implicated in an alternative process of the DNA damage response via regulation of p21 . DDB2 is also involved in SOD2 (superoxide dismutase 2) transcription and stimulation of E2F1-dependent transcriptional targets . DDB2 has been reported to modulate apoptosis due to a complex regulatory circuit between DDB2 and p53 . It has been reported that the ubiquitin-specific protease USP24, a member of the DUBs (deubiquitinating enzymes), cleaves an ubiquitinated form of DDB2 and prevents DDB2 from degradation . A few recent studies also suggest that DDB2 is implicated in cancer development. For example, DDB2 is able to attenuate NF-κB (nuclear factor κB) activity by up-regulating expression of IκB (inhibitor of NF-κB) via binding the proximal promoter of the gene and affect breast cancer progression . It was found that DDB2 is a master regulator of EMT (epithelial–mesenchymal transition) of the colon cancer cells and has a suppressive effect on metastasis of the cancer cells .
The PAQR (progestin and adipoQ receptor) family is a highly conserved family that contains 11 members, PAQR1–PAQR11 . PAQR3 (also known as RKTG for Raf kinase trapping to Golgi) is a seven-transmembrane protein specifically located at the Golgi apparatus in mammalian cells [18,19]. PAQR3 negatively regulates Ras/Raf/MEK/ERK [ERK is extracellular-signal-regulated kinase and MEK is MAPK (mitogen-activated protein kinase)/ERK kinase] and PI3K/Akt (PI3K is phosphoinositide 3-kinase) signalling pathways by sequestering Raf kinase and the PI3K p110α subunit to the Golgi apparatus and, as a result, inhibits the phosphorylation of ERK, Akt and GSK3β (glycogen synthase kinase 3β) [18,20]. PAQR3 functions as a tumour suppressor via its modulatory roles in ERK and PI3K/Akt pathways [21–24]. PAQR3 negatively regulates angiogenesis via modulation of the autocrine function of VEGF (vascular endothelial growth factor) . PAQR3 also plays a vital role in EMT and tumour migration via functional interaction with p53 . In humans, the expression level of PAQR3 was decreased in many types of cancers including colorectal cancer, gastric cancers, osteosarcoma and liver cancers [26–28]. For example, down-regulation of PAQR3 in human gastric cancers is associated with the progression, metastasis and survival in patients with gastric cancers . Collectively, mounting evidence indicates that PAQR3 functions as a tumour suppressor and plays a crucial role in cancer development.
Owing to the paramount importance of PAQR3 in multiple biological functions, it is important to comprehend how PAQR3 itself is regulated. In the present study, we reveal for the first time that DDB2 is implicated in the ubiquitination and degradation of PAQR3. Furthermore, such post-translational regulation of PAQR3 has an impact on the tumour-suppressive activities of the protein. Our results therefore suggest that DDB2 could be involved in the regulation of cancer development via modulation of PAQR3.
MATERIALS AND METHODS
The full-length human PAQR3 and DDB2 cDNAs were cloned from HEK (human embryonic kidney)-293T cells by RT (reverse transcription)–PCR and confirmed by DNA sequencing. The deletion and mutants of PAQR3 were generated by a PCR-based method and confirmed by DNA sequencing. PAQR3 and its mutants were subcloned into the mammalian expression vector pCS2+MT with six Myc tags at the N-terminus. PAQR3 was also cloned into pEGFP-C1 vector to fuse with EGFP at the N-terminus . The full-length and WD domain of DDB2 were subcloned into the mammalian expression vector pRc/CMV with a FLAG epitope tag added to the N-terminus. To generate shRNA-resistant DDB2, the cDNA region containing the targeted sequence of DDB2-specifc shRNA (AGAGCGAGATCCGAGTTTA) was mutated into AAAGTGAAATACGTGTATA by a PCR-based method and confirmed by sequencing. These mutations were chosen without altering the amino acid sequence of DDB2 protein. The shRNA-resistant DDB2 was cloned into pRc/CMV with a FLAG tag added to the N-terminus.
Cell culture and transfection
HEK-293T, HeLa and HepG2 cells were cultured in DMEM (Dulbecco's modified Eagle's medium) containing 10% (v/v) FBS. AGS cells were cultured in Ham's F-12K (Kaighn's) medium containing 10% (v/v) FBS. Cells were maintained at 37°C under 5% CO2 in a humidified incubator. Transient transfection was performed using the PEI (polyethyleneimine) method for HEK-293T cells and PolyJet™ DNA In Vitro Transfection Reagent (SignaGen Laboratories) for HepG2 and HeLa cells.
RNA isolation and quantitative RT–PCR
The cells were lysed in TRIzol® reagent (Invitrogen). Total RNA was purified and reverse-transcribed according to the manufacturer's instructions. Quantitative real-time PCR was conducted with an ABI Prism 7900 sequence detection system (Applied Biosystems) following the manufacturer's recommendations. The primers were as follows: human PAQR3, 5′-CAAGTGCGTCCAGAGAAGATT-3′ and 5′-TCTGACCG-ATGGCAGGAAAAA-3′; human GAPDH (glyceraldehyde-3-phosphate dehydrogenase), 5′-TGCACCACCAACTGCTTAGC-3′ and 5′-GGCATGGACTGTGGTCATGAG-3′; human DDB2, 5′-CTCCTCAATGGAGGGAACAA-3′ and 5′-GTGACCAC-CATTCGGCTACT-3′; human DDB1, 5′-ATGTCGTACAACT-ACGTGGTAAC-3′ and 5-CGAAGTAAAGTGTCCGGTCAC-3′; human CUL4A, 5′-ACCTCGCACAGATGTACCAG-3′ and 5′-AGGTTGACGAACCGCTCATTC-3′; human CUL4B, 5′-ACTCCTCCTTTACAACCCAGG-3′ and 5-TCTTCGCATCAAACCCTACAAAC-3′.
Antibodies, immunoblotting and immunoprecipitation
The antibodies purchased were as follows: mouse anti-FLAG antibody from Sigma; rabbit anti-GM130 (cis-Golgi matrix protein of 130 kDa) antibody from Abcam; anti-phospho-ERK1/2, anti-phospho-Akt (Ser473) and anti-phospho-GSK3β (Ser9) from Cell Signaling Technology; anti-Myc, anti-HA (haemagglutinin), anti-DDB2, anti-total ERK1/2 and anti-tubulin from Santa Cruz Biotechnology; anti-Golgin-97 monoclonal antibody, Alexa Fluor® 488-conjugated donkey anti-mouse IgG, Alexa Fluor® 546-conjugated goat anti-mouse, rabbit IgG and Hoechst 33342 from Invitrogen/Molecular Probes; Cy5 (indodicarbocyanine)-labelled goat anti-mouse and rabbit IgG from Jackson ImmunoResearch; anti-PAQR3 polyclonal antibody from Thermo Fisher Scientific. Immunoprecipitation was performed as described previously . Briefly, cells were washed three times and then lysed for 30 min at 4°C. The homogenates were centrifuged for 20 min at 13400 g/min at 4°C. The supernatant was mixed with antibody and incubated overnight at 4°C. Protein A/G Plus–agarose was added for 4 h at 4°C. The immunoprecipitate was washed three times in cell lysis buffer followed by Western blotting analysis. The protocol for immunoblotting has been described previously .
HeLa cells were grown on glass coverslips. At 48 h after transfection, the cells were fixed with 4% (w/v) paraformaldehyde in PBS, permeabilized with 0.1% Triton X-100 in PBS for 5 min and incubated with primary and secondary antibodies, sequentially. Confocal images were captured using an Olympus FV1200 confocal microscope. The 488 nm line of an argon laser was used for fluorescence excitation of EGFP and Alexa Fluor® 488-conjugated antibody. A helium/neon laser (543 nm) was used for excitation of Alexa Fluor® 546-conjugated antibodies, and 633 nm was used for excitation of Cy5-conjugated antibody. After data acquisition, red/green/blue images were processed using Olympus FV1200 software.
Analyses of protein degradation
At 24 h after transfection, HEK-293T cells were treated with 100 μg/ml cycloheximide and harvested at various time points. The cell lysate was subjected to immunoblotting. For analysis of protein ubiquitination, the cells were co-transfected with Myc–PAQR3 and HA–ubiquitin, together with FLAG–vector or FLAG–DDB2. At 24 h after transfection, the proteasome inhibitor MG132 (10 μM) was added and the cells were incubated for 6 h. To the cell lysates an anti-Myc antibody was added before being incubated overnight at 4°C. Protein A/G Plus–agarose was then added and the mixture was incubated for 4 h at 4°C. The immunoprecipitate was analysed by immunoblotting with anti-HA and anti-Myc antibodies respectively.
Lentivirus and gene silencing by shRNA
Lentivirus-based gene silencing by shRNA was reported previously . In short, the annealed siRNA cassettes with targeting sequences were inserted into the pBS-SKII-hU6 vector downstream of the hU6 promoter. The siRNA expression cassette was then subcloned into the FG12 vector and confirmed by DNA sequencing. The FG12 plasmid containing specific shRNA was directly used in cell transfection in HEK-293T cells to silence expression of endogenous genes. For AGS cells, lentivirus containing specific shRNA was produced and used in viral infection to generate cells with stable expression of the shRNA. The target sequences used in shRNA-related experiments were as follows: AGAGCGAGATCCGAGTTTA for human DDB2; GGACAACCCGTACATCACC for human PAQR3; ACTCAATAAAGTCATCAAA for human DDB1; AGAATATCTTAACCATGTA for human CUL4A; and AATTCTTCAGAAAGGTTTA for human CUL4B.
Assays of cell proliferation
An MTT (Sigma–Aldrich) assay was performed according to the method described by Mosmann . AGS cells were seeded (5×103 cells/well) in a 96-well culture plate and incubated for different lengths of time. At the time of experiment, the culture medium was discarded and the plate was incubated at 37°C for 4 h with MTT (5 mg/ml). The MTT solution was then discarded, and DMSO was added to dissolve the formazan product. The absorbance was measured at 550 nm using a Uniscience spectrophotometer plate reader. For colony-formation assay, AGS cells were cultured on a six-well culture plate (0.5×103 cells per well) and incubated at 37°C. Cells were cultured for 9 days in complete medium containing 10% (v/v) FBS. Then the cells were fixed and stained with 0.005% Crystal Violet. Colonies containing more than 50 cells were counted. All experiments were carried out in triplicate.
Cell migration assays
For Transwell assays, AGS cells were seeded into the upper chamber (8-μm pore size Transwell; Corning Life Sciences) in medium without serum, and medium containing 10% (v/v) FBS in the lower chamber served as attractant. The cells were incubated for 24 h, and cells that migrated to the underside of the membrane were fixed and stained with Crystal Violet, imaged and counted. The wound closure assay was performed as described by Andre et al. . In brief, a wound was induced on the confluent monolayer cells by scraping a gap with a micropipette tip, and the speed of wound closure was monitored by light microscopy. Cell migration was measured by the travelling distance from the original wound edge after 24 h of incubation.
Statistical significance was assessed using Student's t test.
PAQR3 interacts with DDB2 and tethers it to the Golgi apparatus
DDB2 forms an E3 ligase complex with DDB1 and CUL4A and plays a critical part in NER. DDB2 is a WD domain-containing protein and possesses seven WD40 repeats . Our previous study revealed that PAQR3 is able to interact with a WD40 protein, a Gβ subunit of G-proteins . We therefore hypothesized that PAQR3 might also interact with WD-containing DDB2. To test this hypothesis, we used co-immunoprecipitation assays to investigate the protein interactions. We transiently transfected HEK-293T cells with FLAG-tagged DDB2 and Myc-tagged PAQR3. Immunoprecipitation of PAQR3 could pull down DDB2 (Figure 1A). Inversely, immunoprecipitation of DDB2 could pull down PAQR3 (Figure 1B). In addition, we found that PAQR3 could interact with the WD domain of DDB2 (Figure 1C). These data therefore indicate that PAQR3 is able to interact with DDB2 via its WD domain.
Interaction of DDB2 with PAQR3
As PAQR3 is a Golgi-localized protein [18,19], we next investigated whether PAQR3 could alter the subcellular distribution of DDB2. Using immunofluorescence analysis, DDB2 was mainly localized in the nucleus in HeLa cells (Figure 1D, upper panel). This result is consistent with previous reports as DDB2 is mainly involved in DNA damage repair in the nucleus [3–6]. Interestingly, co-expression of PAQR3 profoundly changed the localization of DDB2, leading to most DDB2 being tethered to the Golgi apparatus (Figure 1D, lower panel).
To explore the interaction domain of these two proteins, we constructed six PAQR3 deletion mutants and analysed which deletion(s) could abrogate the protein interaction. Using a co-immunoprecipitation assay, we found that deletions of the amino acid residues of PAQR3 at 1–60 and 40–60 abrogated the ability of PAQR3 to bind DDB2 (Figure 1E), indicating the N-terminal amino acid residues 40–60 of PAQR3 is indispensable for the protein interaction. To confirm further the interaction of PAQR3 with DDB2, we used a GST pull-down assay to analyse whether the N-terminus of PAQR3 could interact directly with DDB2. We found that the GST fusion protein containing the N-terminal 71 amino acid residues could successfully pull down DDB2 protein (Figure 1F). Collectively, these data indicate that the N-terminal end of PAQR3 could interact with DDB2.
Overexpression of DDB2 elevates protein degradation and ubiquitination of PAQR3
What is the functional significance of the interaction between DDB2 and PAQR3? DDB2 is able to form an E3 ubiquitin ligase complex together with DDB1 and CUL4A and has been reported to be involved in protein ubiquitination [7,8]. We hypothesized that DDB2 might be implicated in the ubiquitination and degradation of PAQR3. To test this hypothesis, we first investigated whether the protein level of PAQR3 was affected by DDB2. When HEK-293T cells were overexpressed with DDB2, the endogenous PAQR3 protein level was markedly reduced (Figure 2A). Such a decrease in PAQR3 protein was not due to reduction of PAQR3 gene transcription, as DDB2 overexpression had no effect on the mRNA level of PAQR3 (Figure 2B). Furthermore, we found that overexpression of DDB2 could dose-dependently reduce the protein level of ectopically expressed PAQR3 (Figure 2C). These data therefore provided a hint that DDB2 might modulate the protein stability of PAQR3. Such a notion was supported further by the findings that the half-life of PAQR3 protein in the presence of cycloheximide was significantly reduced by DDB2 overexpression (Figure 2D).
DDB2 regulates degradation of PAQR3 protein
We next analysed whether ubiquitination of PAQR3 was altered by DDB2. HEK-293T cells were transfected with HA-tagged ubiquitin with or without PAQR3 or DDB2. In the presence of MG132, a proteasome inhibitor, PAQR3 underwent a certain degree of protein polyubiquitination (Figure 2E, lane 5). However, such polyubiquitination of PAQR3 was profoundly enhanced by DDB2 overexpression (Figure 2E, lane 6). We confirmed further such a finding by immunofluorescence analysis. We found that when both PAQR3 and DDB2 were co-expressed, they were co-localized with ubiquitin signals (Figure 2F). These data therefore indicate that DDB2 can enhance PAQR3 degradation via a proteasome-dependent pathway by elevating the ubiquitination level of PAQR3.
If PAQR3 were degraded by DDB2, was the function of PAQR3 also affected by DDB2? To address this issue, we elucidated the effect of DDB2 on the modulatory functions of PAQR3 for ERK and PI3K/Akt signalling pathways, as previous studies have demonstrated that these two pathways are regulated by PAQR3 [18,24]. As expected, EGF-induced ERK phosphorylation was reduced by PAQR3 (Figure 2G, lanes 5 and 6 compared with lanes 1 and 2). However, such an inhibitory effect of PAQR3 was abrogated by co-expression of DDB2 (Figure 2G, lanes 7 and 8). On the other hand, insulin-induced phosphorylation of Akt and GSK-3β was decreased by PAQR3 (Figure 2H, lanes 5 and 6 compared with lanes 1 and 2). Such an inhibitory effect of PAQR3 was also abrogated by co-expression of DDB2 (Figure 2H, lanes 7 and 8). These data indicate that DDB2 may affect the biological functions of PAQR3 via altering the protein degradation of PAQR3.
Knockdown of DDB2 elevates the stability and reduces ubiquitination of PAQR3
To explore further the activity of DDB2 in the regulation of PAQR3 degradation, we investigated the functions of DDB2 knockdown on PAQR3. We screened three shRNAs specific for DDB2 and found that one of them could effectively silence the expression of DDB2 (Figure 3A). This shRNA was used in all subsequent experiments. The efficiency of DDB2 knockdown was also confirmed using quantitative RT–PCR (Figure 3B). As expected, the mRNA level of PAQR3 was not affected by DDB2 knockdown (Figure 3B). We found that the protein level of PAQR3 was dose-dependently increased by DDB2 silencing (Figure 3C). Next, we analysed the effect of DDB2 knockdown on the protein stability of PAQR3. In the presence of cycloheximide, the half-life of PAQR3 was markedly elevated by DDB2 knockdown (Figure 3D). At 12–16 h after cycloheximide treatment, the PAQR3 protein level was almost undetectable when DDB2 was not down-regulated. However, PAQR3 protein was barely degraded at these time points when DDB2 was silenced. These data therefore clearly indicate that PAQR3 degradation was affected by DDB2 under physiological conditions.
Knockdown of DDB2 decreases degradation of PAQR3 protein
As DDB2 knockdown could increase the protein level of PAQR3, we postulated that such a change would affect the negative modulatory roles of PAQR3 in cellular signalling. To address this issue, we analysed the activity of DDB2 knockdown in ERK and PI3K/Akt pathways. Consistent with our hypothesis, we found that silencing of DDB2 could reduce EGF-induced ERK phosphorylation (Figure 3E), as well as insulin-induced phosphorylation of Akt and GSK3β (Figure 3F). To ensure that the observed effects were due to reduced DDB2 expression, an shRNA-resistant DDB2 was co-expressed in these cells. Expression of the shRNA-resistant DDB2 could abrogate the inhibitory effect of DDB2 knockdown on EGF- and insulin-mediated signalling (Figures 3E and 3F).
The ubiquitin ligase activity of DDB2 is achieved by a protein complex that contains DDB2, DDB1 and CUL4A [3–6]. To investigate further the function of this complex in ubiquitination of PAQR3, we analysed the effects of silencing DDB1 and CUL4A. As a control, we analysed the function of silencing CUL4B as this protein is not involved in DDB2-mediated protein ubiquitination. We found that knockdown of DDB1 and CUL4A could reduce ubiquitination of PAQR3 (Figure 3G). On the other hand, knockdown of CUL4B had no effect on PAQR3 ubiquitination (Figure 3G). The efficiency of these knockdowns was confirmed by quantitative RT–PCR (Figure 3H). These data therefore confirm that the DDB2–DDB1–CUL4A ubiquitin ligase complex is implicated in the ubiquitination and degradation of PAQR3.
Lys61 of PAQR3 is involved in DDB2-mediated ubiquitination and degradation of PAQR3
We next explored which lysine residue in the PAQR3 protein is implicated in DDB2-mediated ubiquitination. We focused our studies on the N-terminal region of PAQR3 as this domain faces the cytosol and most other regions constitute the transmembrane domains . There are four lysine residues in the N-terminal 71 amino acid residues at positions 4, 7, 41 and 61 in human PAQR3 protein (Figure 4A). We mutated these four lysine residues into arginine individually. As expected, the protein level of wild-type PAQR3 was reduced by DDB2 overexpression (Figure 4B). Intriguingly, DDB2 had no effect at all on the protein level of PAQR3 when Lys61 was mutated (Figure 4B). We next analysed the half-life of PAQR3 protein in the presence of cycloheximide. It appeared that the stability of PAQR3 protein was markedly enhanced when Lys61 was mutated (Figure 4C). Furthermore, we analysed the ubiquitination of these PAQR3 mutants. Mutation of Lys61 led to robust reduction of PAQR3 ubiquitination (Figure 4D). Collectively, these data pinpoint that Lys61 of PAQR3 is involved in DDB2-mediated ubiquitination and degradation of PAQR3.
Lys61 of PAQR3 is involved in DDB2-mediated ubiquitination
DDB2 alters the growth and migration of gastric cancer cells via regulation on PAQR3
Our studies so far have demonstrated that DDB2 is actively involved in the degradation of PAQR3. As PAQR3 functions as a tumour suppressor via its negative modulatory roles on ERK and PI3K/Akt signalling pathways [21–24], we hypothesized that DDB2 might antagonize the tumour suppressive activity of PAQR3 via decreasing the protein level of PAQR3. On the other hand, reducing DDB2 would increase the protein level of PAQR3 and enhance the tumour-suppressor activity of PAQR3. To test this hypothesis, we analysed the effect of silencing DDB2 and PAQR3 on the cellular functions of a gastric cancer cell line AGS as our previous study has revealed that PAQR3 has a suppressive function in human gastric cancers and AGS cells .
We constructed AGS cells that had stable expression of mock vector, DDB2-specific shRNA or PAQR3-specific shRNA. At first, we confirmed the efficiency of silencing DDB2 and PAQR3 in AGS cells by quantitative RT–PCR (Figure 5A). We analysed the cell proliferation rate using an MTT assay. We found that DDB2 knockdown could significantly reduce the cell proliferation rate, whereas such an effect was abrogated by PAQR3 knockdown (Figure 5B). Consistently, knockdown of DDB2 could markedly reduce colony formation of AGS cells and such an inhibitory effect was abrogated by PAQR3 knockdown (Figure 5C). We next analysed the effect of DDB2 on cell migration by Transwell and wound-healing assays. We found that the migration capacity of AGS cell was significantly inhibited by DDB2 knockdown and such an effect was also abrogated by simultaneous knockdown of PAQR3 (Figures 5D and 5E). As expected, we found that DDB2 knockdown could also reduce EGF-stimulated ERK phosphorylation and insulin-stimulated Akt/GSK3β phosphorylation in AGS cells (Figures 5F and 5G). However, such a reduction in EGF- and insulin-mediated signalling was abrogated by simultaneous knockdown of PAQR3 in these cells. Collectively, these data not only indicate that DDB2 could promote cell growth and migration in AGS cells, but also reveal that the effect of DDB2 on cancer cells is mediated by its regulation of PAQR3.
DDB2 antagonizes the tumour-suppressive activity of PAQR3 in gastric cancer cells
Our studies have provided compelling evidence that DDB2 is involved in ubiquitination and degradation of PAQR3. DDB2 is able to interact with PAQR3 in vivo and in vitro. The N-terminal amino acid residues 40–60 of PAQR3 are required for its interaction with DDB2. The WD domain of DDB2 is involved in the interaction between the two proteins. Such an interaction also alters subcellular compartmentalization of DDB2 as PAQR3 tethers DDB2 to the Golgi apparatus. From both overexpression and knockdown experiments, DDB2 is directly involved in the regulation of PAQR3 degradation. The protein expression level, protein stability and ubiquitination of PAQR3 are all changed by the level of DDB2. The regulation of PAQR3 degradation by DDB2 is also physiologically relevant as knockdown of endogenous DDB2 can enhance the protein stability of PAQR3. Furthermore, the modulation of EGF- and insulin-initiated cellular signalling by PAQR3 is altered by DDB2. At the molecular level, we found that Lys61 of PAQR3 is targeted by DDB2 for ubiquitination. Our initial studies also reveal that the DDB1–DDB2–CUL4A ubiquitin ligase complex is involved in the polyubiquitination of PAQR3. Finally, we demonstrate that the regulation of PAQR3 degradation by DDB2 affects the progression of gastric cancer cells. The cell proliferation rate and migration of gastric cancer cells are all inhibited by DDB2 knockdown and such effects are abrogated by PAQR3 knockdown, indicating that the effect of DDB2 on gastric cancer cells is mediated by PAQR3. Collectively, our studies not only pinpoint that DDB2 is a post-translational regulator of PAQR3, but also indicate that DDB2 plays an active role in tumorigenesis via its regulation of PAQR3, a tumour suppressor involved in many types of cancers [25–28].
The functions of DDB2 in tumour progression are complicated, dependent on the molecular mechanism involved and tumour type. The original function of DDB2, as a product of XPE, is involvement in the recognition of UV-damaged DNA and initiation of NER [3–6]. It is therefore postulated that DDB2 functions as a caretaker in the repair of UV-damaged DNA. Under this scenario, DDB2 can be considered as a tumour suppressor in cancers related to UV-induced DNA damage. This notion is supported by the findings that deletion of the Ddb2 gene in mouse causes skin tumours and cells deficient in DDB2 are resistant to apoptosis induced by UV light but not by chemical carcinogens . In another study, it was found that DDB2-depleted mice developed spontaneous tumours . These findings using mouse models are consistent with human conditions in which defective activity of the DDB2-containing complex causes a repair defect in patients with XPE that is an autosomal recessive disorder characterized by photosensitivity and early onset of carcinomas. In addition, p53 and its target gene p21 are involved in the tumour-suppressive activity of DDB2 in cancer cells. It was reported that melanoma cells acquire resistance to DNA cross-linking agents via p53-dependent up-regulation of DDB2 . DDB2 could also co-operate with p21 to suppress UV-induced skin malignancies . Collectively, these findings indicate that DDB2 may function as a tumour suppressor via its activities in DNA repair or response to DNA damage mainly in skin cancers.
DDB2 may also participate in cancer formation due to transcriptional regulation of genes critical for carcinogenesis. DDB2 was reported to negatively regulate the transcription of the SOD2 gene and impact on breast cancer growth . Overexpression of DDB2 inhibited the motility, invasiveness and metastasis of human breast cancer cells . Such a negative effect of DDB2 on tumour invasion was caused by negative regulation of NF-κB activity by DDB2 via transcriptional regulation of IκB through direct binding of DDB2 to the promoter of the IκB gene . DDB2 was also reported to be a master regulator of EMT, a critical event for cancer invasion, in colon cancer . The effect of DDB2 on EMT was found to be linked to transcriptional regulation of VEGF, Zeb1 and Snail by DDB2 .
In addition to the involvement in cancer development via modulating DNA repair and gene transcription, DDB2 can regulate cancer progression through interaction with proteins critical in cancer cells. In the DDB1–DDB2–CUL4A ubiquitin ligase complex, DDB2 serves as a substrate receptor that determines the specificity of targeted substrate . It can be postulated that the substrate targeted by the DDB1–DDB2–CUL4A complex is not limited to proteins involved in NER. It was reported that DDB2 can interact with AR (androgen receptor) and target AR for ubiquitination and degradation . Consistently, DDB2 could inhibit cell growth in AR-expressing cancer cells, but not in AR-null cells . DDB2 could also interact with Artemis, a member of the SNM1 gene family, and then target a CDK (cyclin-dependent kinase) regulator p27 for protein degradation . In the present paper, we report that PAQR3 is another protein targeted by DDB2. We found that DDB2 antagonizes the tumour-suppressive activity of PAQR3 in gastric cancer cells. The cell proliferation rate and migration of gastric cancer cells are inhibited by DDB2 knockdown. As these effects of DDB2 knockdown are abrogated by PAQR3 knockdown, the tumour-promoting activity of DDB2 on gastric cancer cells is mediated by its regulation of PAQR3. We propose that DDB2 may play a vital role in cancer progression in the types of tumour where PAQR3 has a prominent tumour-suppressive activity. Overexpression of DDB2 in these types of tumour would abrogate the functions of PAQR3, thereby promoting tumour progression. In addition, we predict that DDB2 may target other proteins for degradation and affect many biological processes, including those involved in carcinogenesis. Future studies in these directions will not only broaden our knowledge about the multifaceted functions of DDB2, but also provide new strategies for the treatment of human diseases such as malignancies.
Shanshan Qiao and Yan Chen designed the experiments and wrote the paper. Shanshan Qiao performed the experiments. Weiwei Guo, Lujian Liao, Lin Wang, Zheng Wang, Rui Zhang, Daqian Xu, Yuxue Zhang, Yi Pan and Zhenzhen Wang provided technical assistance.
This work was supported by research grants from Ministry of Science and Technology of China [grant numbers 2012CB524900 (to Y.C.) and 2013BAI04B03 (to Z.W.)] and National Natural Science Foundation of China [grant numbers 81130077, 81390350 and 81321062 (to Y.C.)].
damage-specific DNA-binding protein
Dulbecco’s modified Eagle’s medium
cis-Golgi matrix protein of 130 kDa
glycogen synthase kinase 3β
human embryonic kidney
inhibitor of NF-κB
nucleotide excision repair
nuclear factor κB
progestin and adipoQ receptor
superoxide dismutase 2
vascular endothelial growth factor
xeroderma pigmentosum group C
xeroderma pigmentosum group E