Targeting UBR5 inhibits postsurgical breast cancer lung metastases by inducing CDC73 and p53 mediated apoptosis

Abstract

UBR5 is a HECT domain E3 ubiquitin ligase that is frequently amplified in breast, ovarian, and prostate cancers. Heightened UBR5 expression plays a profound role in tumor growth through immune-dependent mechanisms; however, its mode of action in driving tumor metastasis has not been definitively delineated. Herein, we used a tetracycline (Tet)-inducible RNAi-mediated expression silencing cell system to investigate how UBR5 enables postsurgical mammary tumor metastatic growth in mouse lungs without the continuous influence of the primary lesion. In vitro, Ubr5 knockdown induces morphological and molecular changes characteristic of epithelial-mesenchymal transition (EMT). In vivo, UBR5 promotes lung metastasis in an E3 ubiquitin ligase-dependent manner. Moreover, doxycycline-induced UBR5 expression knockdown in metastatic cells in the lungs, following removing the primary tumors, resulted in increased apoptosis, decreased proliferation, and prolonged survival, whereas silencing the expression of cell division cycle 73 (CDC73), a tumor suppressor and E3 ligase substrate of UBR5, reversed these effects. Transcriptome analyses revealed a prominent role of the p53 pathway in dovitinib-induced apoptosis of tumor cells differentially regulated by UBR5 and CDC73. In human triple-negative breast cancer (TNBC) patient specimens, a strong inverse correlation was observed between UBR5 and CDC73 protein levels, with reduced CDC73 expression at metastatic sites compared to primary lesions. Furthermore, a xenograft model of human TNBC recapitulated the metastatic properties and characteristics of the unique UBR5-CDC73 functional antagonism. This study reveals the novel and critical roles and intricate relationships of UBR5, CDC73, and p53 in postsurgical breast cancer metastasis and indicates the potential of targeting this pathway in cancer therapy.

Graphical Abstract

Introduction

Breast cancer metastasis is the leading cause of cancer-related mortality, accounting for approximately 90% of mortalities¹. Typically, triple-negative breast cancer (TNBC) has a high frequency of visceral metastases and poor prognosis, with few therapeutic options. Moreover, majority of studies have focused on primary lesions, and effective therapeutic options are very limited for distant metastatic breast cancer owing to the lack of specific targets. Therefore, there is an urgent need to identify critical oncogenic targets that are responsible for TNBC invasion and distant expansion and to develop effective targeted therapeutic agents to halt post-surgery metastatic recurrence in TNBC patients.

Tumor metastasis is the product of a process involving multiple phases. Most circulating tumor cells (CTCs) survive dissemination and seeding. However, only approximately 0.02% of them can colonize and develop into macro-metastases². Mesenchymal-to-epithelial transition (MET), wherein cells with mesenchymal characteristics are converted into epithelial cells, enables disseminated tumor cells (DTCs) survival and colonization in the lung parenchyma³. WNT signaling is essential for metastasis-initiating fitness in lung cancer cells. Breast cancer cells educate fibroblasts to produce tenascin C, which subsequently activates WNT signaling and enhances metastatic colonization⁴.

Ubiquitin protein ligase E3 component n-recognin 5 (UBR5, also known as EDD), originally identified as a progesterone-induced gene in human breast cancer using differential display^5–7, has been reported to be amplified in several malignant tumors, such as breast and ovarian cancers⁶. It is a HECT domain E3 ubiquitin ligase that destabilizes proteins via N-terminal recognition. We previously identified a profound role of UBR5 in promoting experimental TNBC growth, predominantly through the adaptive immune system mediated by CD8⁺ T cells. However, the metastasis-promoting property of UBR5 appears to be “cell-intrinsic”^{8, 9}. UBR5 has been demonstrated to interact directly or indirectly with several proteins involved in the regulation of cancer cell metabolism, proliferation, and apoptosis^{10, 11}. However, the molecular mechanisms by which UBR5 promotes cancer metastasis, including its E3 ubiquitin ligase activity, remain obscure.

Herein, we evaluated the role of UBR5 in TNBC lung metastasis using an inducible gene expression knockdown experimental system to manipulate UBR5 levels in metastatic sites following surgical removal of the primary tumor, evaluated the colonization of disseminated tumor cells in the lungs at the cellular and molecular levels, and explored the underlying cellular and molecular mechanisms.

Materials and Methods

Cell lines and cultures

All cell lines were acquired from ATCC and authenticated using short tandem repeat (STR) profiling within the last 3 years. All experiments were conducted using mycoplasma-free cells. 4T1 (RRID:CVCL_0125) and MDA-MB-231 (RRID: CVCL_0062) cells were cultured in the appropriate medium supplemented with 10% fetal bovine serum. To establish stable 4T1/GFP cells, the cells were transfected with TurboGFP (SHC003; Sigma) and selected using puromycin. To generate 4T1/Ubr5^−/−+ hUBR5 and 4T1/Ubr5^−/−+C2768A cell lines, pCMV-Tag2B UBR5 and pCMV-Tag2B C2768A (Addgene, #37188 and #37189, respectively) were transfected into 4T1/Ubr5^−/− cells using Lipofectamine 3000 reagent (L3000008, Invitrogen, Waltham, MA). Stable cell lines were selected using G418.

Plasmid, lentivirus preparation, and transduction

To construct pLKO-Tet-shUbr5, pLKO-tet-scrambled, and pLKO-Tet-shCdc73 plasmids, Ubr5-shRNA (sequences are listed in Supplemental Table 1) and scrambled shRNA Cdc73-shRNA were cloned into the Tet-pLKO-puro (Addgene, plasmid # 21915) and Tet-pLKO-neo vector (Addgene, plasmid #21916). To construct the pLKO-shUBR5 and pLKO-shCDC73 plasmid, UBR5-shRNA and CDC73-shRNA was cloned into the pLKO.1-puro (Addgene, plasmid #8453) and pLKO.1-neo vector (Addgene, plasmid #13425).

Lentiviral plasmids were mixed with psPAX2 (Addgene, plasmid 12260) and pMD2.G (Addgene, plasmid 12259) at a 4:3:1 ratio. 4T1 and MDA-MB-231 cells were transduced with lentivirus using 10ug/ml polybrene (Santa Cruz Biotechnologies) and selected with puromycin or G418.

Tumor models

Female BALB/c mice aged 6–8 weeks and 8-week-old female NOD.Cg-Prkdc<scid> Il2rg<tm1Wjl>/SzJ (NOD Science Gamma or NSG) mice were purchased from Jackson Laboratory. 5×10⁵ 4T1 and 5×10⁶ MDA-MB-231 cells were injected orthotopically into the BALB/c and NSG mice abdominal mammary glands. BALB/c mice bearing 4T1 tumors were anesthetized with isoflurane, and the tumors were resected. Mice were euthanized at the indicated time points for analysis, and survival was monitored. The tumor volume was computed using the equation (L × W²)/2, where “L” =length and “W” = width.

Quantification of the number of macro- and micro-metastases in the lungs

The number of overt macro-metastases on the lungs surface was enumerated manually. The lungs were perfused with Indian ink via the trachea, removed, and destained in Feketes solution. To quantify micrometastases, whole lungs were excised and digested as previously descripted⁸ and single cells were cultured in medium comprising 60 mol/L 6-thioguanine. Tumor colonies were stained with crystal violet and enumerated after 10 days selection.

Total RNA preparation, quantitative real-time PCR, and RNA-seq

Total RNAs was purified using the RNeasy Mini Kit (QIAGEN). 74106) and an RNase-Free DNase Set (QIAGEN. 79254). RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, 4368814). Quantitative real-time PCR was conducted using the PowerUp SYBR Green reagent (Thermo Fisher Scientific). The relative amount of target mRNA was normalized to that of the endogenous control (GAPDH). (All primer sequences are listed in Supplementary Table 2)

Second-generation RNA sequencing was conducted at the Genomic Core Facilities of Weill Cornell Medicine. The libraries were prepared using the Illumina Stranded mRNA Sample Library Preparation kit (Illumina, San Diego, CA, USA) and were sequenced with paired-end 50 bps on a NovaSeq6000 sequencer. The RNA reads were aligned to the mouse reference genome (GRCm38) using STAR (Version2.5.2). The abundance of transcripts was measured with Cufflinks in FPKM and differential expression with DEseq2 was conducted¹². Gene set enrichment analysis (GSEA) was used to identify pathways and gene sets. A |NES|>1 and p value ≤ 0.01 was considered statistically significant. All statistical analyses were conducted using R4.2.1.(The sequencing coverage and quality statistics for each sample are summarized in Supplementary Table 3.)

Western Blotting

Cells were lysed in RIPA buffer as previously descripted⁸. Protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to PVDF membranes. The membranes were immunoblotted with primary antibodies: UBR5 (sc-515494, Santa Cruz), Parafibromin (A300–170A, Bethyl Laboratories), PARP (9542, Cell Signaling Technology), p53 (ab26, Abcam), GAPDH (sc-FL335, Santa Cruz), and ꞵ-actin (sc-47778, Santa Cruz). Detection and quantification of band intensities were conducted using the ImageJ software. Western blotting was conducted in duplicates.

Clonogenicity assay

Tumor cells were seeded in 6-well plates (200 cells/well) and cultured at 37 °C in a humidified atmosphere of 5% CO₂. After 10 d, the cells were stained with 0.2% crystal violet, and the number of colonies formed in each well was counted and photographed under a microscope.

Wound healing and transwell migration assays

The wound healing assay was conducted by seeding cells in 6-well plates until they reached 95% confluence, as previously described. Transwell assays were conducted using 24-well plates (354578; Corning). Starved tumor cells were seeded into the upper Boyden chambers⁸. Images were captured using a digital camera attached to the inverted microscope at the indicated time points. The images were evaluated and quantified using Image software.

Cell proliferation assay

Cell proliferation was measured using sulforhodamine B (SRB) assay. Cells were seeded in 96-well plates, cultured for 72 h, fixed with 10% trichloroacetic acid (Sigma, T8657), and stained with 0.4% (w/v) SRB (Sigma, 230162) in 1% acetic acid solution. The absorbance (OD) was measured at 510 nm.

Histology, immunohistochemical staining, and TUNEL staining

Tissue microarrays of human TNBC specimens (BR1301, BR1901, and BR1902) were purchased from US Biomax, Inc. Animal specimens (tumors and lungs) were collected, fixed in 10% buffered formalin, and embedded in paraffin. Primary antibodies against UBR5 (NB100–1591, Novus), parafibromin (sc-33638, Santa Cruz Biotechnology), Ki67 (ab16667, Abcam), E-Cadherin (NBP2–19051SS, Novus Biologicals) and Vimentin (NBP1–31327, Novus Biologicals) were used for immunohistochemical analyses. After deparaffinization, rehydration, and antigen retrieval, the slides were incubated with 0.3% H₂O₂ and blocked (ab64226), followed by incubation with primary antibody (4 °C, overnight) and goat anti-rabbit HRP secondary antibody. Slices were observed and images were captured using an Olympus BX60 Upright Microscope. A tissue apoptosis assay was conducted using the TUNEL assay kit (Promega, G7360).

Immunohistochemical staining (IHC) analysis was conducted by two independent pathologists who were blinded to the sample information. Each section was randomly selected from five fields at 400Ｘ magnification. UBR5, CDC73, Ki-67, E-cadherin and Vimentin staining were quantified using multiple intensities, and positive cell percentage scores were determined as previously described¹³. The staining intensity (SI) was recorded as follows: 0, none; 1, weak; 2, moderate; and 3, strong. The percentage of positive cells (PP) was recorded as follows: 0, none; 1, <10% positive; 2, 10%–50%; 3, 51%–80%; and 4, >80%. The semi-quantitative immunoreactive score (IRS) was computed as PP× SI.

Bioinformatic analysis of UBR5 and CDC73 mRNA expression in breast cancer specimens

Bioinformatic analysis of breast cancer specimens was conducted using data published in cBioPortal (https://www.cbioportal.org/). UBR5 mRNA expression data were retrieved and evaluated¹⁴.

Flow cytometry and cell apoptosis analyses

Single-cell suspensions were prepared as previously descripted⁹. After live dead staining with Zombie UV TM Fixable Stain, all samples were stained with the appropriate surface antibodies, including CD3 (17A2), CD8α (53–6.7), B220 (RA3–6B2), CD11b (M1/70), CD11c (N418), F4/80 (BM8), Ly6 G-APCcy7 (Clone 1A8), Ly6 C-PEcy7 (Clone HK1.4), and MHC II (Clone CD11 c (N418) from BioLegend and CD45 (30-F11) and CD4 (GK1.5) from eBioscience. All antibodies were tested using an isotype control. Apoptosis was evaluated using the FITC Annexin V Apoptosis Detection Kit I (catalog no.556547; BD Biosciences). Data acquisition was conducted using FACSymphony A5 and analyzed using FlowJo 10.7.2.

Protein expression, purification, and assay for E2 discharge from E2-Ub by UBR5 E3

Protein UBR5 and UBR5 C2768A expression and purification were conducted as previously described¹⁵. Proteins were purified using a Superdex 200 column. The N-terminal 6xHis-tag on Ub was cleaved and purified using a Superdex 75 column (GE Healthcare).

Purified Ub was loaded onto purified E2. Fluorescently labeled Ub was incubated with E1 and E2D2 to produce Ub-charged E2 (Ub–E2D2). Ub–E2D2 was incubated with purified wild-type or mutant UBR5. The reactions were run on SDS-PAGE gels, and the band intensities were quantified using the ImageJ software.

Oxygen Consumption Rate (OCR) analysis

Oxygen consumption rate (OCR) was measured with The Cell Mito Stress Test Kit (Seahorse, 103015–100) by using Seahorse Analyzer XFe96. 20,000 4T1 cells were seeded in XF96 cell culture microplate and allowed to attach overnight. Prior to the assay, cells were prewashed 3 times with XF assay medium supplemented with 10 mM glucose, 1 mM pyruvate and 2 mM L-Glutamine and then incubated at 37°C without CO2 for 60 minutes. The OCR was measured as cells were treated sequentially with oligomycin (1.5 μM), FCCP (1.0 μM) and rotenone and antimycin-A (0.5 μM).

Statistical analysis

All data points and bar graphs represent the mean of independent biological triplicates. Data are depicted as the mean ± SEM. All statistical analyses were conducted using the GraphPad Prism 9 software. Unpaired two-sided Student’s t-test was used to compare two groups of data; one-way or two-way ANOVA was conducted for comparisons of multiple conditions; the log-rank test was used to compare the survival rates; and Pearson’s correlation coefficient was utilized to evaluate the relationship between UBR5 and CDC73 expression in TNBC patient samples.

Results

Ubr5 silencing impairs primary tumor growth and lung metastasis

To evaluate UBR5 function in breast cancer lung metastasis in a temporal and dynamic manner, we established a cell system. Herein the pLKO lentiviral vector expressing a Ubr5-specific shRNA was stably integrated into the mouse TNBC cell line 4T1, henceforth referred to as “Tet-shUbr5 cells,” for inducible endogenous Ubr5 expression knockdown via the addition of doxycycline (Dox). Using this cell-based system, we demonstrated that reducing UBR5 expression enhanced the migratory and invasive capacities of cancer cells by inducing EMT-like alterations, whereas their ability to form colonies is reduced (Supplementary Fig. S1).

GFP or Tet-shUbr5 4T1 tumor cells were inoculated into the mammary pad, and the mice were administered different amounts of Dox daily by gavage from day 3 onwards after tumor cell inoculation to induce Ubr5 expression knockdown (Fig 1A and Fig. S2). We found that 100 mg/kg Dox significantly inhibited primary tumor growth, as measured by both tumor volume (Fig 1B) and weight (Fig 1C), due to Ubr5 expression knockdown in vivo, which is consistent with our previous study⁸. The knockdown of UBR5 expression in the primary tumor at the mRNA and protein levels was evaluated by RT-PCR, western blotting, and tissue IHC staining (Fig 1D–G). Furthermore, Ubr5-knockdown in tumor cells caused a significant reduction in lung colonization (Fig 1H). These data demonstrate that UBR5 is crucial for 4T1 tumor growth and metastatic lung lesion formation.

Figure 1. Limited primary tumor growth and less lung metastasis in mice carrying inducible Ubr5-knockdown 4T1 cells.

(A) Schematic diagram of the experimental design. 5×10⁵ 4T1-GFP, 4T1-Tet-shUbr5 cells are injected into the mammary gland of female BALB/c mice and on D3, the mice are treated with 100 mg/kg Dox by gavage daily. On d 27, mice are sacrificed, and tumors are dissected. (B) 4T1 tumor growth curve. (C) Tumor weight is measured. (D) Ubr5 mRNA expression of mice bearing tumor is detected by real-time quantitative PCR. (E). Protein level of UBR5 in tumor analyzed by Western blot. (F) Representative UBR5 IHC staining of tumor from mice bearing 4T1/GFP, 4T1/Tet-shUbr5 cells. (G) Statistical analysis of UBR5 immunostaining IRS score. (H) Palpable metastatic nodules on lung surfaces are enumerated. Data represents mean ± SEM; **P <0.01; ***P < 0.001.

Silencing Ubr5 suppresses post-surgery lung metastasis via its E3 ubiquitin-ligase

Although Ubr5 knockdown resulted in increased mobility and invasiveness of 4T1 tumor cells, the newly arrived cells failed to progress to macroscopic lung metastases. To decipher this apparent paradox in a way that would eliminate the interference of the effects of the primary tumor on metastases in the lung, we removed the primary tumor surgically on D20, followed by daily Dox administration from D23 onward to induce the knockdown of Ubr5 expression in tumor cells that had already reached the lung before surgery (Fig 2A and Fig. S3). On D35, the mice were sacrificed and lung metastases were measured by staining the lungs with Indian ink (Fig 2B, C) and conducting the 6-thioguanie clonogenicity assay (Fig 2D, E). Both assays revealed strong reductions in lung metastases following Dox administration, which was associated with the extended survival of the mice (Fig 2F).

Figure 2. Inducible Ubr5 expression knockdown suppresses post-surgery lung metastasis.

(A) Schematic for removing primary tumor and knocking down Ubr5 expression with daily Dox addition (100 mg/kg) by gavage post-surgery in mice. (B) and (C) Representative images for lung nodules stained with India ink (B) Lung nodules in tumor-bearing are quantified (C). (D) and (E) The 6-thioguanine clonogenicity assay of Tet-scr and Tet-shUbr5 tumors (on D35) for lung metastasis is conducted. Images of representative experiment are depicted (D), and colonies are quantified (E). (F) Overall survival rates of the mice (Tet-scr, n=8; Tet-scr(+Dox), n=6; Tet-shUbr5, n=5; Tet-shUbr5(+Dox), n=5). (G) UBR5 protein expression in hUBR5- and C2768A-reconstituted 4T1/Ubr5^−/− cells is evaluated in Western blotting. (H) Representative result of E2 discharge assay for UBR5 and C2768A. (I) Quantitation of the discharged E2-Ub by UBR5 and C2768A. (J) A total of 5×10⁵ Tet-scr, Tet-shUbr5, Ubr5^−/−+hUBR5, or Ubr5^−/−+C2768A cells are injected into 6- to 8-week-old female BALB/c mice (with 8, 8, 8, and 16 mice, respectively, for the four cell lines). Tumor volume is measured on D20. (K) The metastatic cells in the lung are enumerated by the 6-thiogaunine assay. (L) After removing the primary tumor, the mice with similar primary tumor volumes are selected from all groups and administrated with Dox daily by gavage (100 mg/kg). On D30, the mice are sacrificed. Lung metastatic nodules are stained with Indian ink and quantified (Tet-scr, n=5; Tet-shUbr5, n=5; Ubr5^−/−+hUBR5, n=5, Ubr5^−/−+C2768A, n=4). Data represents mean ± SEM; *P <0.05; **P <0.01; ***P < 0.001.

To further investigate the mechanism of UBR5 metastasis-promoting function in the lung, we reconstituted the 4T1/Ubr5^−/− cells that we previously generated via CRISPR⁸ with hUBR5, which shares 98.3% identity at the amino acid level with murine UBR5, or the E3 ubiquitin ligase-deficient mutant C2678A (Fig 2G). We verified the loss of C2678A E3 ubiquitin ligase activity using a classical E2 discharge assay (Fig 2H, I). In vivo, C2768A-reconstituted Ubr5^−/− 4T1 cells displayed a significantly slower primary tumor growth rate by D20 than the other three groups that had normal UBR5 levels (Fig 2J), suggesting that the effect of UBR5 on primary tumor growth is partially dependent on its E3 ligase activity. We selected mice from the four groups that had primary tumors of similar sizes, surgically removed them, and treated them with Dox. At this time point, the four groups of mice had similar numbers of micrometastases in the lungs before Dox administration (Fig 2K and Fig. S3). Notably, Dox-induced Ubr5 knockdown greatly diminished lung metastatic nodules, whereas hUBR5 expression rescued metastasis to levels observed in the WT control. However, C2768A mutant expression in tumor cells failed to sustain lung metastatic nodules (Fig 2L). These data suggest that UBR5 exhibits primary lesion-independent pro-metastatic tumor activity in the lungs in a manner that is completely dependent on E3 ubiquitin ligase activity.

Silencing Ubr5 expression in the lung causes metastatic tumor cell apoptosis

To elucidate the molecular mechanism of action of UBR5 in breast cancer lung metastasis, we examined immune cells in the pulmonary microenvironment. However, we did not find significant different (Fig. S4). UBR5 suppresses apoptosis in several malignant tumor cells¹⁶. We evaluated metastatic 4T1 tumor cells apoptosis in the lungs using the TUNEL assay. From D28 onward, five days after the mice were treated with Dox, we observed that the lungs of mice bearing Dox-induced Ubr5-knocked metastatic tumor cells exhibited persistently increased TUNEL-positive signals compared to the control groups (Fig 3A). Conversely, decreased Ki67 expression was observed in Ubr5-knockdown lung nodules (Fig 3B). These data demonstrate that metastatic tumor cells in the lungs undergo increased apoptosis and decreased proliferation following DOX-induced knockdown of Ubr5 expression.

Figure 3. Silencing Ubr5 expression causes apoptosis in 4T1 tumor cells in the lung.

(A) Representative images of TUNEL staining at indicated time post tumor removal surgery of mice with injection of Tet-scr and Tet-shUbr5 cells (n=3 mice per group), Scale bar: 100 μm. (B) Ki67 expression in metastatic nodules in the lung. (C) Dose response of 4T1 cells treated with Dovitinib for 48 h with respect to cell viability. (D) The protein level of PARP-1, Cleaved(c)-PARP and (E) Proportion of apoptosis cells of Tet-scr and Tet-shUbr5 cells treated with 1μm Dovitinib for 24 h. Data represents mean ± SEM; **P <0.01; ***P < 0.001.

UBR5 has been reported to contribute to chemoresistance by suppressing MYC-mediated apoptotic priming in breast cancer cells¹⁷. We observed that Dox-induced Ubr5 knockdown significantly reduced 4T1 cells resistance to dovitinib-induced cell death, a potent multi-target tyrosine kinase (RTK) inhibitor, with a strongly reduced IC50 (Fig 3C). Consistent with this observation, caspase-mediated poly-ADP ribose polymerase 1 (PARP-1) cleavage was also significantly enhanced following Dox-induced Ubr5 expression knockdown in the presence of dovitinib (Fig 3D). This observation was corroborated by FACS analysis using propidium iodide (PI) and annexin V staining (Fig 3E). These data support the strong and localized role of UBR5 in suppressing apoptosis and promoting the proliferation of metastatic lung tumor cells in a post-surgery setting.

UBR5 suppresses tumor apoptosis and promotes metastatic growth by targeting CDC73

We recently identified the tumor suppressor parafibromin, encoded by CDC73, as a substrate of E3 ubiquitin ligase activity of UBR5 and a target that is degraded via the proteasome in TNBC cells¹⁸. Herein, we reaffirmed that Dox-induced Ubr5 knockdown in 4T1 cells (Fig S5A) does not affect Cdc73 mRNA expression (Fig S5B) but affects the protein level (Fig S5C). Moreover, CDC73 targeting by UBR5 was dependent on E3 ligase activity (Fig S5D).

To evaluate the role of CDC73 in UBR5-regulated apoptosis and proliferation, we transiently knocked down the expression of Ubr5, Cdc73, or both in 4T1 tumor cells using Tet-shRNA (Fig 4A). Dox-induced Ubr5 knockdown did not affect spontaneous apoptosis but resulted in a marked increase in PI/Annexin V-positive apoptosis induced by dovitinib, which was completely reversed by further Cdc73 expression knockdown (Fig 4B, C). These results indicated that the apoptosis and chemosensitivity of 4T1 tumor cells regulated by UBR5 were opposite to those regulated by CDC73.

Figure 4. UBR5 suppresses apoptosis and promotes lung metastasis via targeting CDC73.

(A) UBR5 and CDC73 protein expression in 4T1 cells with Dox-induced shUbr5 and shCdc73 expression. (B) Representative FACS images of Dox-treated Tet-scr, Tet-shUbr5, Tet-shCdc73 and Tet-shUbr5-shCdc73 cells exposed to Dovitinib (1μm) for 24 h. (C) Quantitation of apoptotic cells in b. (D) Apoptosis in lung sections of mice bearing Dox-treated Tet-scr, Tet-shUbr5, Tet-shCdc73 or Tet-shUbr5-shCdc73 tumors, respectively, is evaluated by TUNEL staining on D35 (n = 3 mice per group), Scale bar: 100 μm. (E) Representative IHC staining for Ki67 expression images in the lung with quantifications. (F) Lung metastatic nodules in tumor-bearing are stained with India ink and quantified (5 mice/group). (G) Overall survival rates of mice carrying Dox-treated Tet-scr (n=7), Tet-shUbr5 (n=5), Tet-shCdc73 (n=6), or Tet-shUbr5-shCdc73 (n=8) tumors. Data represents mean ± SEM; *P <0.05; **P <0.01; ***P < 0.001.

In vivo, Dox-induced Ubr5 knockdown increased tumor cell apoptosis in the lungs, as revealed by TUNEL staining, and this effect was completely reversed by reducing Cdc73 expression (Fig 4D). Conversely, DOX-induced knockdown of Ubr5 expression decreased the number of Ki67⁺ proliferating cancer cells in the lungs, which was reversed by reducing CDC73 levels (Fig 4E). Furthermore, DOX-induced Ubr5 knockdown decreased the number of metastatic lung nodules (Fig 4F) and significantly extended the survival of the mice, which was reversed by reducing Cdc73 expression (Fig 4G).

Collectively, these data demonstrate that UBR5 promotes breast cancer lung metastasis and survival through its E3 ubiquitin-ligase activity to target the CDC73 protein, which antagonizes UBR5’s effects on both apoptosis and proliferation.

UBR5 knockdown activates p53 pathway via CDC73

We took further steps to comprehend the molecular mechanisms underlying the regulatory activities of UBR5 and CDC73 in dovitinib-induced apoptosis via RNA-seq analysis in 4T1 tumor cells that were Dox-silenced for the expression of Ubr5, Cdc73, or both and treated with dovitinib. Gene set enrichment analysis (GSEA) of the RNA-seq data querying the hallmark gene sets specified in the Molecular Signatures Database (MSigDB) revealed significant positive enrichment of several key pathways, including Wnt_beta-catenin signaling, p53 activation, and inflammation, in shUbr5-treated vs. control 4T1 cells (Fig 5A). However, these apoptosis-related activities were significantly reversed in shUbr5-treated cells, and further silencing of Cdc73 expression accentuated the enrichment of the G2M-checkpoint, E2F targets, and Notch signaling-related and pro-proliferative activities (Fig 5B). Consistent with the theme of apoptosis, we found that the apoptotic pathway was strongly enriched in the Ubr5 knockdown subgroup (Fig 5C), whereas this enrichment was reversed by further Cdc73 expression knockdown (Fig 5D). Notably, the p53 signaling pathway displayed responses similar to those of the apoptotic pathway (Fig 5E, F). Consistent with this observation, QT-PCR and Western blot analyses demonstrated that Ubr5 knockdown increased p53 mRNA and protein levels, but this effect was reversed by silencing Cdc73 expression (Fig 5G, H).

Figure 5. UBR5 and CDC73 regulate apoptosis via p53 pathway.

4T1 cells are treated with Dox and Dovitinib (1μm, 24 h) and RNA are extracted for RNA-seq analysis. (A), (B) Gene set enrichment analysis (GSEA) results with hallmark gene sets in 4T1 Tet/shUbr5 vs Tet/scr cells (A) and Tet/shUbr5-shCdc73 vs Tet/scr (B). (C)-(F) GSEA plots of gene sets of interests are displayed for Apoptosis in Tet/shUbr5 vs Tet/scr cells (C) and in Tet/shUbr5-Cdc73 vs Tet/shUbr5 cells (D) and p53 pathway signaling in Tet/shUbr5 vs Tet/scr cells (E) and in Tet/shUbr5-Cdc73 vs Tet/shUbr5 cells (F). (G) QT-PCR analysis of Trp53 mRNA level. (H) Western blot analysis of the p53 protein level in the four groups of cells. (I) Heat map of differentially expressed genes involved in the p53 signaling pathway from the four groups of 4T1 cells treated with Dox and Dovitinib. (J) The protein level of UBR5 and p53 in 4T1 cells following siRNA-mediated Ubr5 expression silencing. (K) Representative FACS images and quantitation of apoptotic cells of WT, WT-siTp53, KO, KO-siTp53 cells exposed to Dovitinib (1μm) for 24 h. Data represents mean ± SEM; **P <0.01; ***P < 0.001.

Upon further examination of the p53 pathway genes by hierarchical analysis, the GSEA heatmap revealed an interesting group of genes whose mRNA expression was strongly induced in shUBR5-treated cells (indicated by a blue box in Fig 5I) but was suppressed to the levels of the scrambled control group in Ubr5 and Cdc73 double-knocked down cells, including Trp53 itself. This pattern of expression was consistent with that of the dovitinib response in the four groups of cells regulated by UBR5 and CDC73, suggesting that targeting UBR5 in tumor cells promotes drug-induced apoptosis via CDC73, and that UBR5 controls p53 expression at the mRNA level through CDC73.

To further determine the role of p53 in UBR5-controlled apoptosis, we knocked down Tp53 expression via siRNA in WT and Ubr5^−/− 4T1 cells in the presence of dovitinib (Fig 5J) and analyzed the changes in apoptosis by FACS (Fig 5K). The data shows that knocking down the elevated p53 level in Ubr5^−/− cells largely but not completely reduced the apoptosis induced by dovitinib, indicating a major albeit partial dependence on p53.

UBR5 and CDC73 play crucial roles in human TNBC metastasis

To ascertain clinical evidence for the role of UBR5 and CDC73 in human breast cancer, we evaluated UBR5 and CDC73 protein levels using IHC staining in a cohort of 70 primary TNBC tumor tissue sections. A significant negative correlation was observed between the UBR5 and CDC73 protein levels (Fig 6A, 6B). Moreover, we evaluated CDC73 mRNA levels using publicly available human patient-derived data from cBioportal and found that its levels were lower in metastatic lesions than in primary tumors (Fig 6C). To investigate the function of UBR5 in human TNBC metastasis, we knocked down UBR5 expression in MDA-MB-231 cells, which resulted in a marked increase in CDC73 protein levels (Fig S6A). Furthermore, similar to the 4T1 cells, UBR5 knockdown increased the migratory capacity of MDA-MB-231 cells (Fig S6B, C).

Figure 6. Effects of human TNBC-expressed UBR5 on tumor growth and lung metastases.

(A) IHC staining of UBR5 and CDC73 proteins in primary tumors from TNBC patients (n=70 samples). Scale bars: 200 μm. (B) Correlation between tumor UBR5 and CDC73 expression levels in human TNBC patients. (C) mRNA levels of CDC73 in primary breast cancer tissues (n=168) and metastatic sites (n=52). (D) Metastatic nodules in the lung and liver of tumor-bearing mice were quantified on D45. (E) Representative IHC staining of UBR5 and CDC73 in lung metastatic nodules. Scale bar: 100 μm. (F) UBR5 and CDC73 protein expression in MDA-MB-231-Vector, MDA-MB-231-shUBR5 or MDA-MB-231-shUBR5-shCDC73 cell lines. (G) Tumor growth rates of NSG mice. (H) Metastatic nodules in the lung and liver of tumor-bearing mice were quantified on D38. Data represents mean ± SEM; *P <0.05; **P <0.01; ***P < 0.001.

To further evaluate the functional relationship between UBR5 and CDC73 in lung metastasis, WT and UBR5 knockdown MDA-MB-231 cells were implanted into the NSG mice mammary glands. The primary tumors exhibited little difference in growth rates (Fig S6D), consistent with our previous report of an immune-dependent mechanism of UBR5-mediated tumor growth⁹. However, UBR5 knockdown in tumors caused a strong reduction in metastatic nodules in the lungs and liver compared with the WT (Fig 6D). Moreover, UBR5 knockdown strongly upregulated CDC73 expression in lung metastatic nodules (Fig 6E), echoing the human clinical observations described above (Fig 6C). To further explore the role of CDC73 in UBR5-dependent lung metastasis, we knocked down CDC73 in UBR5-silenced cells (Fig 6F). Upon inoculation of NSG mice, the rate of primary tumor growth was similar, regardless of the UBR5 or CDC73 levels (Fig 6G). However, the knockdown of CDC73 expression in UBR5-deficient cells restored lung and liver metastases to WT levels (Fig 6H). Collectively, these results suggest that tumor derived UBR5 is crucial for human breast cancer metastasis, and CDC73 opposes UBR5’s metastatic activities.

Discussion

Despite major breakthroughs in breast cancer treatment in recent years, little is known about the treatment response of metastatic cells after they disseminate to secondary organs, and distant recurrence remains a grave clinical challenge¹⁹. Herein, we aimed to determine how the oncological protein UBR5 acts in the lung microenvironment that encompasses metastatic breast cancer implantation, survival, colonization, and progression in a postsurgical model of TNBC, wherein UBR5 was silenced in an inducible manner, allowing a clear view of the effects of UBR5 on metastatic cells alone, independent of UBR5 activity in the primary tumor.

This study revealed for the first time that tumor derived UBR5 is required for post-surgery metastatic tumor growth and is responsible for poor survival in patients with TNBC. Diminishing UBR5 expression in both mouse 4T1 and human MDA-MB-231 cancer cells after removing the primary tumor impairs metastatic tumor growth at secondary sites in a manner that is completely dependent on UBR5 E3 ligase activity and CDC73-mediated apoptosis, independent of the adaptive immune system. This study provides a mechanistic rationale for targeting UBR5 as a novel and potentially efficacious therapeutic approach for the post-surgical metastasis of TNBC.

Paradoxically, UBR5 knockdown in breast cancer cell lines induces greater invasive and migratory capacities in vitro, whereas it results in reduced colony growth in vitro and lung metastatic nodules in vivo. In this context, it is noteworthy that, compared to the WT control, UBR5-deficient tumor cells exhibited increased expression of PDGFRβ, whose signaling in carcinomas is well established with respect to angiogenesis²⁰. PDGFRβ has been demonstrated to promotes the invasion and migration of mammary tumors by regulating EMT via the PI3K/AKT/mTOR pathway^21–23. Conversely, Lama3, Areg, and Itga2 were downregulated in UBR5-deficient cells. LAMA3 overexpression has been reported to downregulate the expression of invasion-related markers (MMP2 and MMP9) and EMT-related proteins (N-cadherin and vimentin) in lung adenocarcinoma cells²⁴. Amphiregulin (AREG) is one of several ligands capable of binding and activating the epidermal growth factor receptor (EGFR)^{25, 26} and accelerating the MET process during reprogramming²⁷. ITGA2 is consistently overexpressed in several cancers and is thought to be involved in cell adhesion^28–30. Moreover, irreversibly switching off UBR5 expression in metastatic tumor cells in the lungs “fixed” them in the mesenchymal phenotype, evidenced by decreased E-cadherin expression and increased Vimentin expression (Fig. S7). Collectively, our data suggest that, although the non-reversible loss of UBR5 expression in tumor cells promotes EMT, it impairs the MET process and its ability to colonize distant organs and form metastatic lesions.

We previously demonstrated that UBR5 promotes primary mammary tumor growth in a cell-extrinsic manner through cytokine and chemokine production, which affects CD8⁺ T cell-mediated immunity via paracrine actions^{9, 31}. Our study provides definitive evidence that metastatic tumor growth driven by UBR5 is cell-intrinsic, since UBR5-deficient MDA-MB-231 cells, similar to WT cells, are unable to survive in the lungs of NSG mice. To note, UBR5 was recently reported to promote the mitochondrial oxidative phosphorylation (OXPHOS) process via upregulating the expression of proline-rich tyrosine kinase 2 (PYK2) in colorectal carcinoma cell lines³² and OXPHOS plays an important role in cancer metastasis³³. However, we did not observe a significant difference in mitochondrial oxygen consumption rate (OCR) between WT and Ubr5^−/− 4T1 cells (Fig. S8).

Apoptosis is a type of programmed cell death which can eliminate “out of control” tumor cells. Thus, a high apoptosis rate serves as a constraint for malignant tumor metastasis³⁴. p53 is a tumor suppressor that induces apoptosis and inhibits metastasis by upregulating several pro-apoptotic genes^{35, 36}. UBR5 has been demonstrated to directly interact with numerous proteins implicated in a vast variety of cellular processes, including the cell cycle, DNA damage and repair, apoptosis, cell adhesion, and transcription⁶. Breast cancer cell lines that are less resistant to cell death have lower UBR5 expression, and UBR5 depletion increases sensitivity to Fas-ligand-mediated apoptosis³⁷. We observed increased apoptosis and decreased proliferation of Ubr5 knockdown lung metastatic nodules in the 4T1 model.

CDC73 is a UBR5-mediated ubiquitination substrate¹⁸. This study demonstrated that impaired lung colonization of metastatic cells due to UBR5 deficiency could be completely reversed by reducing CDC73 expression. It indicates that CDC73 upregulation caused by UBR5 deficiency significantly accounts for decreased metastatic tumor growth in the lungs and shortened animal survival. Mice with a deleted Cdc73 allele in the parathyroid glands displayed increased proliferation compared to control mice with dual alleles and developed parathyroid and uterine neoplasms³⁸. CDC73 induces growth arrest and promotes the initiation and/or execution of apoptosis resulting from DNA damage and other cytotoxic stimuli^{39, 40}. Cdc73 knockdown reversed dovitinib-induced apoptosis and proliferative deficits in UBR5-deficient tumors in vivo. GSEA analysis of the transcriptomes revealed prominent enrichment of the apoptosis-pathway in Ubr5 knockdown cells, rendering them more sensitive to drug-induced apoptosis mediated by CDC73. Interestingly, one of the most significantly affected biological processes in UBR5 vs. CDC73 antagonism was the p53 pathway, which is well known to be essential for apoptosis induction in out-of-control tumor cells. This pathway was strongly altered by the loss of Ubr5 expression, with marked Trp53 expression upregulation at the mRNA level, which was significantly reversed by further silencing of Cdc73 expression. Smits first reported that UBR5 downregulation results in elevated p53 protein levels in both transformed and untransformed cells, triggering a senescent phenotype in fibroblasts⁴¹. UBR5 actively inhibits p53 phosphorylation by ATM and plays a role in G(1)/S progression in breast cancer cells⁴². We have previously reported that in ovarian cancer cells, UBR5 is essential to sustain cell-intrinsic-catenin-mediated signaling to promote cellular adhesion/colonization and organoid formation by controlling p53 protein level³¹.

CDC73 is a PAF protein complex component, which can function as a tumor suppressor or oncoprotein in a context-dependent manner⁴³. Tyrosine-phosphorylated CDC73 forms a complex with SUV39H1 and mediates the transrepression of Wnt target genes such as cyclin D1 and c-myc. However, upon tyrosine dephosphorylation by SHP2, CDC73 stably binds-catenin. The dephosphorylated CDC73/β-catenin interaction overrides CDC73/SUV39H1-mediated transrepression⁴⁴. Moreover, Zheng et al. demonstrated that forced CDC73 expression in human colorectal cancer cells promotes cyclin B1, D1, E, p38 MAPK, and p53 expression⁴⁵. Collectively, these data suggests that UBR5-controlled CDC73 promotes p53 transcription, thereby activating the p53 pathway that induces apoptosis.

In summary, this study revealed a novel and important activity of UBR5 with cell-intrinsic metastasis-promoting properties by resisting tumor cell apoptosis and facilitating their survival and propagation in the lungs, which is specifically antagonized by CDC73 via p53 in a regulatory signaling pathway. This study highlights the crucial role of UBR5 in TNBC colonization at secondary sites, independent of the adaptive immune system, and highlights the therapeutic potential of targeting UBR5 and/or its signaling pathways to improve the prognosis of patients with TNBC with dissemination and chemoresistance.

Novelty and Impact:

UBR5 plays a profound role in promoting tumor growth. However, its mode of action in driving tumor metastasis has not been definitively delineated. Using a tetracycline (Tet)-inducible system to knock down UBR5 in postsurgical mice, wherein the primary mammary tumor was removed, this study revealed the novel roles of UBR5, CDC73, and p53 in breast cancer metastasis by regulating the disseminated cancer cells apoptosis and implicates the potential of targeting this pathway in cancer therapy.

Abbreviations:

CDC73

Cell division cycle 73

CTCs

Circulating tumor cells

Dox

Doxycycline

DTCs

Disseminated tumor cells

EMT

Epithelial-mesenchymal transition

ER

Estrogen receptor

GSEA

Gene set enrichment analysis

HPT-JT

Hyperparathyroidism-jaw tumor

MET

Mesenchymal to epithelial transition

OCR

Oxygen consumption rate

OXPHOS

Oxidative phosphorylation

PARP-1

Poly-ADP ribose polymerase 1

PC

Parathyroid cancer

PDGFRβ

Platelet-derived growth factor receptor β

PR

Progesterone receptor

RTK

Potent multi-targeted tyrosine kinase

TNBC

Triple-negative breast cancer

UBR5

Ubiquitin protein ligase E3 component n-recognin 5

Data Availability Statement

The RNA-seq data generated in this study are available in GEO under accession number GSE222787. Publicly available mRNA expression data of CDC73 in human breast cancer were acquired from the Metastatic Breast Cancer (MSK, Cancer Discovery 2022), The Metastatic Breast Cancer Project (Archived, 2020), Metastatic Breast Cancer (INSERM, PLoS Med 2016) and The Metastatic Breast Cancer Project (Provisional, December 2021) studies published in cBioPortal (https://www.cbioportal.org/study/summary?id=breast_ink4_msk_2021%2Cbrca_mbcproject_wagle_2017%2Cbrca_igr_2015%2Cbrca_mbcproject_2022). Other data supporting the findings of this study are available from the corresponding author upon request.