Ubiquitin C-terminal hydrolase L1 is a regulator of tumor growth and metastasis in double-negative prostate cancer

Abstract

Prostate cancer is the second leading cause of cancer-related deaths among men worldwide. With heavy androgen deprivation therapies, prostate cancer may shift to androgen receptor negative and neuroendocrine negative subtype of castration resistant prostate cancer, defined as double-negative prostate cancer. Double-negative prostate cancer is associated with poor prognosis and disease mortality. The molecular mechanisms underlying the emergence of double-negative prostate cancer remain poorly understood. Here, we demonstrate that Ubiquitin C-Terminal Hydrolase L1 (UCH-L1), is negatively correlated with androgen receptor levels in prostate cancer patients. UCH-L1 plays a functional role in tumorigenesis and metastasis in double-negative prostate cancer. Knock-down of UCH-L1 decreases double-negative prostate cancer colony formation in vitro and tumor growth in vivo. Moreover, decrease of UCH-L1 significantly delays cell migration in vitro and spontaneous metastasis and metastatic colonization in vivo. Proteomic analysis revealed that mTORC1 signaling, androgen response signaling and MYC targets are the top three decreased pathways upon UCH-L1 decrease. Further, treatment with LDN-57444, a UCH-L1 small molecule inhibitor, impairs double-negative prostate cancer cell colony formation, migration in vitro, and metastatic colonization in vivo. Our study reveals that UCH-L1 is an important regulator of double-negative prostate cancer tumor growth and progression, providing a promising therapeutic target for this subtype of metastatic prostate cancer.

Introduction

Prostate cancer remains the most frequently diagnosed non-cutaneous cancer type in men and the second leading cause of cancer-associated death in men [1]. Metastatic prostate cancer is often more challenging to treat and is associated with a significant decrease in survival rates and poor clinical prognosis [1-3]. Prostate cancer progression is facilitated by different molecular mechanisms, including alterations of AR signaling, genomic alterations, epigenetic modifications, metabolic reprogramming, lineage plasticity, and activation of aberrant signal pathways [3-47]. Identification of new drivers of prostate cancer tumorigenesis and progression is crucial for developing new therapeutic strategies aimed at preventing and treating metastatic prostate cancer.

The standard of care for patients presenting with metastatic prostate cancer is androgen-deprivation therapies (ADT) [48-53]. With hormone therapy, metastatic prostate cancer eventually develops treatment resistant, referred as castration resistant prostate cancer (CRPC) [46]. Metastatic CRPC can be further defined into five subtypes based on androgen receptor (AR) and neuroendocrine (NE) markers status including AR+/NE- adenocarcinoma, AR+/NE+ double-positive, AR^low/NE- low AR, AR-/NE- double-negative, and AR-/NE+ neuroendocrine prostate cancer (NEPC) [4,11,19,54-56]. Double-negative prostate cancer presents distinct molecular and clinical features with the shortest survival rate [29,57-61]. It was demonstrated that double-negative prostate cancer is caused by AR ablation and bypasses AR dependence by activating the MAPK/FGF pathway [57]. Nevertheless, the molecular drivers and therapeutic targets for double-negative prostate cancer remain elusive.

Ubiquitin C-Terminal Hydrolase L1 (UCH-L1) has a dual function in regulating protein stability and degradation with critical roles in regulating various cellular processes, including protein degradation and stability, apoptosis, and cell cycle regulation [62-66]. UCH-L1 can act as an oncogene as well as a tumor suppressor in cancer [67-70]. In the context of prostate cancer, UCH-L1 is silenced by promoter methylation in localized prostate cancer, while it is re-expressed and highly upregulated in 56% of NEPC [70,71]. However, the functional role of UCH-L1 in double-negative prostate cancer remains unknown.

Here, we report that UCH-L1 expression negatively correlates with AR expression. Further, we demonstrate that UCH-L1 regulates double-negative prostate cancer growth and metastasis. Modulation of UCH-L1 decreases mTOR signaling, androgen response signaling, and MYC targets signaling which provides potential molecular mechanisms underlying UCH-L1 regulation of double-negative prostate cancer. More importantly, treatment with a UCH-L1 inhibitor, LDN-57444, decreases double-negative prostate cancer cell growth and migration in vitro and metastatic colonization in vivo. Our findings suggest that UCH-L1 is a potential molecular indicator and therapeutic target for double-negative prostate cancer.

Results

UCH-L1 regulates cell growth and tumor growth of double-negative prostate cancer in vitro and in vivo

We first analyzed the correlation of UCH-L1 mRNA levels with AR mRNA levels in published patient datasets [23,72,73]. Our data demonstrated that UCH-L1 mRNA levels were negatively correlates with AR mRNA levels in three independent prostate cancer cohorts (Grasso, et al.: R = -0.35, P = 0.0006; Kumar, et al.: R = -0.44, P < 0.00001, and Taylor, et al.: R = -0.28, P = 0.0013) (Figure 1A). These data suggest that high expression of UCH-L1 may contribute to AR loss during disease progression to AR negative prostate cancer. To test the functional role of UCH-L1 in double-negative prostate cancer, we silenced UCH-L1 in a double-negative prostate cancer cell line, DU-145, by two independent shRNAs. The knock-down efficiency of UCH-L1 was tested by Western Blot (Figure 1B). Downregulation of UCH-L1 in DU-145 significantly reduced colony formation ability compared to control (Figure 1C). Likewise, silencing UCH-L1 in DU-145 decreased subcutaneous tumor growth in vivo (Figure 1D). Decease of UCH-L1 also significantly decreased proliferation rate measured by Ki67 positive cells in the tumors (Figure 1E, 1F and Supplementary Figure 1). These data suggest that UCH-L1 regulates double-negative prostate cancer cell growth and tumorigenesis in vitro and in vivo.

Figure 1.

UCH-L1 regulates double-negative prostate cancer cell growth and tumor growth. A. mRNA levels of UCH-L1 and AR were obtained from three independent datasets. The correlation of UCH-L1 and AR mRNA levels were analyzed. R and p values were indicated. B. Western Blot analysis of UCH-L1 levels in DU-145-shControl (shCtl), DU-145-shUCH-L1 #1 (sh#1), and DU-145-shUCH-L1 #2 (sh#2) cells. C. Colony formation assays of DU-145-shCtl, DU-145-shUCH-L1 #1 and #2 cells. The percentage of colony area per well was quantified using ImageJ. Scale bars = 1 cm. D. Subcutaneous tumor growth of DU-145-shCtl, DU-145-shUCH-L1 #1, and #2 cells. Tumor weights at the end point were shown on the right graph. Error bars represent standard error of the mean (SEM). E, F. Immunohistochemistry (IHC) staining for UCH-L1 and Ki67 in DU-145-shCtl, DU-145-shUCH-L1 #1 and #2 xenografts. Scale bars represent 50 microns. Percentage of Ki67 positive cells of DU-145-shCtl, DU-145-shUCH-L1 #1 or DU-145-shUCH-L1 #2 tumors were quantified. Error bars represent SD. For all, *P < 0.05, **P < 0.01, ***P < 0.005, and n.s. = not significant, determined by Student’s t-test (two-tailed) at each time-point.

UCH-L1 regulates cell migration in vitro and metastasis in vivo

We next tested UCH-L1 function on double-negative prostate cancer metastasis in vitro and in vivo. Migration assay and wound healing assay revealed that knockdown of UCH-L1 in DU-145 cells significantly reduced migratory abilities in vitro (Figure 2A, 2B). These results support a role of UCH-L1 in driving an aggressive phenotype in prostate cancer.

Figure 2.

UCH-L1 regulates double-negative prostate cancer migration and invasion in vitro and in vivo. A. Migratory abilities of DU-145-shCtl, DU-145-shUCH-L1 #1, and DU-145-shUCH-L1 #2 cells were assessed by Transwell chamber assays. Experiments were performed in triplicate and five images per well were used for quantification. The percentage of migration were normalized to the DU-145-shCtl cell counts. Representative images of one of three independent experiments were shown. Scale bars indicate 100 microns. B. Wound healing assays of DU-145-shCtl, DU-145-shUCH-L1 #1, and DU-145-shUCHL1 #2 cells. Whole well pictures were scanned with Celigo at 0, 24, and 48 hours. The average of the wound area in three wells at 0 hour was assigned 100% and used to quantify changes in wound area over time. C. Bioluminescence imaging of DU-145-shCtl and DU-145-shUCH-L1 spontaneous model (n = 7). Whole-body bioluminescence intensity was quantified at 3 weeks post-surgery (bottom panel). D. Percentage of the mice with metastases was quantified by red fluorescence signals on each organs including lungs, lymph node (LN), liver, kidneys, and bone. E. Fluorescence images of lungs from DU-145-shCtl and DU-145-shUCH-L1. Scale bars represent 2 mm. F. RFP fluorescence images of liver, lymph node, and kidney from DU-145-shCtl and DU-145-shUCH-L1. Scale bars represent 2 mm. For all, *P < 0.05, **P < 0.01, and ***P < 0.005, determined by two-tailed Student’s t test.

We further tested UCH-L1 function on metastasis using subcutaneous xenograft tumor model of spontaneous metastasis (Figure 2C-F). DU-145-shUCH-L1 and control (shCtl) cells were transduced with lentivirus expressing red fluorescent protein (RFP) and luciferase (Luc) for quantification and visualization of metastasis nodules. Quantification of whole-body bioluminescence signals and incidences of individual organs showed a significant decrease in DU-145-shUCH-L1-RFP-Luc group when compared to DU-145-shCtl-RFP-Luc group (Figure 2C, 2D). Additionally, compared with DU-145-shCtl-RFP-Luc group, both the size and the number of the metastatic nodules of lungs displayed smaller in the UCH-L1 knockdown group (Figure 2E). It is notable that all DU-145-shCtl-RFP-Luc (7/7 mice) had lymph node metastasis, while only 4/7 mice of DU-145-shUCH-L1-RFP-Luc group were found metastases in lymph nodes (Figure 2F).

Knockdown of UCH-L1 in DU-145 cells decreases metastatic colonization in vivo

To further test the role of UCH-L1 in double-negative prostate cancer metastasis, we adopted another well-established intracardiac injection metastasis model to test the role of UCH-L1 on the ability of DU-145 cells to home at distant sites (Figure 3) [70,74,75]. Similarly, bioluminescence imaging and quantification revealed that silencing of UCH-L1 in DU-145-RFP-Luc cells significantly decreased prostate cancer metastatic colonization when compared to DU-145-shCtl-RFP-Luc cells (Figure 3A). By quantifying red fluorescence signals in diverse organs, including liver, lungs, kidneys, and bone, we further demonstrated a remarkable decrease in the metastasis incidences, numbers, and size of metastatic nodules upon UCH-L1 knockdown in double-negative prostate cancer in comparison of the control group (Figure 3B-F). Importantly, bone is the most frequent metastatic site of prostate cancer, and no metastasis was found in bone from the DU-145-shUCH-L1-RFP Luc (0/8 mice) group (Figure 3E), suggesting knockdown of UCH-L1 diminishes metastatic colonization in double-negative prostate cancer. We confirmed that efficiency of UCH-L1 knockdown in metastatic nodules in liver, lungs, and kidneys by immunochemistry (IHC) staining (Figure 3G). Taken together, our results demonstrate that downregulation of UCH-L1 decreases double-negative prostate cancer metastasis, suggesting that targeting UCH-L1 could be a potential therapeutic strategy for suppressing double-negative prostate cancer metastatic progression.

Figure 3.

Downregulation of UCH-L1 decreases double-negative prostate cancer metastatic colonization. (A) DU-145-RFP-Luc-shCtl (n = 7) and DU-145-RFP-Luc-shUCH-L1 (n = 8) cells expressing luciferase and RFP were injected into the left ventricle of male NSG mice. Whole-body bioluminescence intensity (photons/second/cm²/surface radiance) was quantified at Day 21. (B) Percentage of the mice with metastases was quantified in organs including liver, lung, kidney, and bone based on fluorescence signals. (C-F) Representative images of metastatic nodules in liver, lungs, bone, and kidneys of DU-145-shCtl and DU-145-shUCH-L1 were detected with fluorescent stereomicroscopy. Number and size of liver (C), lung (D) and kidney (F) metastases based on counting of the RFP foci using ImageJ. (G) IHC staining of UCH-L1 in liver, lung, and kidney from DU-145-shCtl and DU-145-shUCH-L1 (scale bar = 50 microns). *P < 0.05 and **P < 0.01, determined by two-tailed Student’s t test.

UCH-L1 regulates mTORC1 signaling, androgen response signaling, and MYC targets in double-negative prostate cancer

UCH-L1 has a dual function in regulating protein degradation and stability [76,77]. To identify potential mechanism of how UCH-L1 regulates double-negative prostate cancer growth, we conducted proteomic analysis of two independent DU-145-UCH-L1 knockdown xenografts compared to DU-145-shCtl xenografts. 100 proteins were significantly decreased and 93 proteins were increased upon UCH-L1 knockdown when compared to control xenografts with cut off p value < 0.001 and fold change < -1.5 or > 1.5 (Log2) (Figure 4A, 4B). Pathway analysis demonstrated that mTORC1 signaling, androgen response, and MYC targets signaling were the top three decreased pathways in DU-145-shUCH-L1 xenografts when compared to control xenografts (Figure 4C). Knockdown of UCH-L1 decreases molecular functional pathways associated with RNA binding and nucleosomal DNA binding (Figure 4C, 4D). Our results provide potential molecular mechanisms underlying UCH-L1 modulates double-negative prostate cancer tumor growth and metastasis.

Figure 4.

UCH-L1 regulates mTORC1 signaling, androgen response signaling and MYC targets in double-negative prostate cancer. A. Volcano plot of the significantly increased (red) and decreased (blue) proteins. Cut-off is based on p value < 0.001 and fold change > 1.5 or < -1.5 (Log2). B. Heatmap displaying fold change of the significantly increased and decreased proteins in DU-145-shUCH-L1 xenografts compared to DU-145-shCtl xenografts. C. Gene set enrichment analysis (GSEA) of decreased proteins upon UCH-L1 knockdown from MSigDB Hallmark 2020 and GO Molecular Function 2023. D. Functional protein association networks of decreased protein upon UCH-L1 knockdown in DU-145 xenografts were analyzed using STRING (https://stringdb.org/). The purple node indicates a cluster of proteins related to RNA binding. The blue node indicates a cluster of proteins related to nucleosomal DNA binding. Line thickness indicates the strength of data that support the correlation between each node.

Inhibition of UCH-L1 decreases cell growth, migration, invasion, and metastatic colonization

Lastly, we tested the therapeutic effect of UCH-L1 inhibition on double-negative prostate cancer cell growth, migration, and invasion in vitro and metastatic colonization in vivo. Treatment with UCH-L1 inhibitor, LDN-57444, significantly decreased cell growth and reduced migration and invasion abilities measured by colony formation, transwell migration, and invasion chamber assays (Figure 5A-C). Based on these findings, we further tested UCH-L1 inhibition on metastasis accessed by intracardiac injection model. Although no significant reduction of whole-body bioluminescence signals was found in LDN-57444 treated group compared to vehicle group in DU-145-RFP-Luc model (Figure 5D, 5E), we observed a significant decrease in lung metastasis incidences (3/10 mice) when compared to vehicle group (8/8 mice) (Figure 5E). By quantifying the red fluorescence signals, we observed less numbers of metastases in lungs and liver from LDN-57444 treated group relative to vehicle group (Figure 5F). Altogether, our results demonstrate that UCH-L1 inhibition could be a potent therapeutic strategy for targeting double-negative prostate cancer progression.

Figure 5.

Inhibition of UCH-L1 decreases colony formation, migration, and invasion in vitro and in vivo. A. Colony formation assays of DU-145 cells treated with LDN-57444 (0, 5, and 10 uM). The percentage of colony area per well was quantified using ImageJ and was plotted as percentage of colony formation relative to vehicle group. Scale bars = 1 cm. B. Transwell chamber assays of DU-145 cells treated with LDN-57444 (2.5 uM) or vehicle. Scale bars = 100 microns. C. Invasion assay of DU-145 cells treated with LDN-57444 (2.5 uM) or vehicle. Scale bars represent 100 microns. D. Bioluminescence imaging of intracardiac injection metastasis model of the DU-145-RFP-Luc cells treated with LDN-57444 (5 mg/kg, corn oil, i.p., daily) or vehicle on day 14 post-treatment. The bioluminescence signal was quantified by fold change compared with day 0. E. Percentage and number of metastasis-positive animals/total animal number by organ site. F. Representative fluorescence images of metastasis organs including bone, lung, and liver. The number of metastasis nodules were quantified based on red fluorescence signals in lungs and liver (right panel). Scale bar = 2 mm. For all, *P < 0.05, and n.s. = not significant, determined by Student’s t-test.

Discussion

Double-negative prostate cancer, defined by AR negative and NE negative, accounts for 21-25% of castration resistant prostate cancer and is associated with worst clinical outcome in prostate cancer [29,54,57,78]. Understanding the molecular mechanisms underlying double-negative prostate cancer initiation and progression is essential for the development of effective and molecular based targeted therapy for this lethal disease. Previous studies demonstrated that activation of FGFR and MAPK signaling pathways promotes double-negative prostate cancer growth which is independent of AR activity [57,79]. Another study demonstrated that PRC1 is highly expressed in double-negative prostate cancer specifically and regulates double-negative prostate cancer stemness and immune suppression [80]. A combination of targeting PRC1 with immunotherapy serves as a potent therapeutic strategy for double-negative prostate cancer [80]. PRRX2, a target of the TGF-β pathway, was also identified as a key regulator in double-negative prostate cancer by CRISPR activation screening [81]. The PRRX2 gene signature can stratify the double-negative prostate cancer population and is associated with poor clinical outcomes [81]. Based on histologically characterizing of double-negative prostate cancer models, LuCaP PDX models, double-negative prostate cancer presents with features of squamous cell carcinoma [54].

UCH-L1 is crucial for the progression and metastasis of multiple cancer types including lymphoma, breast cancer, and neuroendocrine carcinomas [69,70,82,83]. Our study suggests that the mechanisms underlying the role of UCH-L1 in double-negative prostate cancer may involve alterations in protein stability and signaling pathways that are involved in mTOR signaling, androgen response signaling, and MYC targets. Our study identifies a new key regulator for double-negative prostate cancer, suggesting that UCH-L1 is a novel therapeutic target for this lethal disease. With limited therapeutic benefit observed with LDN-57444, future study can test a new UCH-L1 specific inhibitor, IMP-1710 [84], on double-negative prostate cancer tumor growth and metastasis.

Collectively, we demonstrate that UCH-L1 mRNA levels are negatively associated with AR mRNA levels in prostate cancer patient cohorts which provides clinical evidence of the importance of UCH-L1 in AR-low or AR-negative prostate cancer. Our study demonstrates that silencing of UCH-L1 decreases double-negative prostate cancer cell growth, migration, and invasion in vitro, and spontaneous metastasis and colonization at distant sites in vivo. Moreover, inhibition of UCH-L1 decreases cell growth, migration, and metastatic colonization in double-negative prostate cancer. Our study demonstrates that UCH-L1 could be a key regulator of double-negative prostate cancer tumorigenesis and progression, which provides a potential therapeutic target for metastatic prostate cancer.

Targeting UCH-L1 expression and activity could offer a novel therapeutic approach for the treatment

Materials/subjects and methods

Clinical datasets

All mRNA levels were obtained from datasets on cBioPortal (www.cbioportal.org/). UCH-L1 mRNA z-score and AR mRNA z-score were obtained and analyzed from Grasso, et al. Nature, 2012; Kumar, et al. Nat Med 2016; and Taylor, et al. Cancer Cell, 2010 [23,72,73]. The correlation of UCH-L1 and AR mRNA levels were calculated by Prism 10.

Cell line and cell culture

DU-145 cells were purchased from the American Type Culture Collection (Cat#HTB-81, ATCC; Manassas, VA). All cells were maintained in RPMI medium supplemented with 10% FBS, 4 mM L-glutamine, and 1% penicillin/streptomycin.

Generation of control and knock-down cell lines

FUCRW lentiviral plasmid was a kind gift from Dr. Owen Witte’s laboratory at University of California, Los Angeles. The pHIV-LucZsGreen plasmid is from Addgene (gift from Bryan Welm, Addgene plasmid # 39196; http://n2t.net/addgene:39196; RRID: Addgene_39196). The DU-145 cells expressing luciferase and red fluorescent protein (RFP) were generated as described previously [74]. DU-145-RFP-Luc were infected with lentiviruses carrying shControl (shCtl) RNA and two independent UCH-L1 shRNAs. The control shRNA is from Addgene (gift from David Sabatini, Addgene plasmid # 1864; http://n2t.net/addgene:1864; RRID: Addgene_1864) (CCTAAGGTTAAGTCGCCCTCGCTCGAGCGAGGGCGACTTAACCTTAGG). Validated small hairpin RNA targeting UCHL1 sequences were purchased from Millipore Sigma (St. Louis, MO). shUCH-L1 #1: TRCN0000007274; Sequence: 5’CCGGCGGGTAGATGACAAGGTGAATCTCGAGATTCACCTTGTCATCTACCCGTTTTT-3’; shUCH-L1 #2: TRCN0000007276; Sequence: 5’CCGGCCAGCATGAGAACTTCAGGAACTCGAGTTCCTGAAGTTCTCATGCTGGTTTTT-3’. Cells were selected in puromycin (0.5 µg/mL) for 9 days.

Colony formation assay

For generation of knockdown cells, five hundred DU-145 (shCtl, shUCH-L1#1, shUCH-L1#2) cells were plated per well in 6-well plates. Colonies were cultured for nine days with medium was refreshed every third day. For drug treatment, 5×10² DU-145 cells were seeded per well in 6-well plates. The cells were treated with vehicle or LDN-57444 (5 or 10 uM) for nine days and the medium with compound was refreshed every three days. After nine days, colonies were fixed with ice cold methanol for 30 minutes and stained with 0.01% crystal violet for 20 minutes. Plates were scanned and quantified based on the percentage of colony-covered area per well. All experiments were performed in triplicate with three biologic replicates, and presented as mean ± SD.

Migration and invasion assay

For knocking down cells, 5×10⁴ DU-145-shCtl, shUCH-L1#1, or shUCH-L1#2 cells were seeded in 300 uL serum-free medium in 24-well transwell inserts (Transwell™ Permeable Polyester Membrane Inserts). The inserts were incubated in 800 uL medium with 10% FBS in 24-well plates for 22 hours. For drug treatment, DU-145 cells were pre-treated with LDN-57444 (2.5 uM) or vehicle for 48 hours. Pre-treated DU-145 cells (5×10⁴) were seeded in 300 uL serum-free medium with or without LDN-57444 (2.5 uM) in 24-well transwell inserts (migration assay) or Matrigel-coated Boyden chamber PET membrane (invasion assay). The inserts were incubated in 800 uL medium with 10% FBS and with or without LDN-57444 (2.5 uM) in 24-well plates for 22 hours. The cells that migrated or invaded the membrane to the bottom were fixed and stained with 0.01% crystal violet and manually counted. Migrated or invaded cells were normalized based on controls (shCtl or vehicle) and the percentage of migration and invasion were calculated. All experiments were performed in triplicate with three biological replicates, and presented as mean ± SD.

Wound healing assay

DU-145-shCtl, shUCH-L1#1, or shUCH-L1#2 (5×10⁶) cells were seeded into 6-well plates. After cells attached and achieved full confluency, one linear scratch per well was generated using a 200 ul pipette tip. The plates were scanned by Celigo at 0, 24, and 48 hours after scratch, and migrative area was quantified at 24- and 48-hour using ImageJ software. All experiments were performed in triplicate with three biological replicates and presented as mean ± SD.

Western blot

DU-145-shCtl, shUCH-L1#1, or shUCH-L1#2 cells were collected by cell scraper and centrifuged at 500 g for three minutes. The cells were lysed in RIPA buffer with protease and phosphatase inhibitors (1:100, Thermo Fisher Scientific) for 30 mins. 40 ug proteins from DU-145-shCtl, shUCH-L1#1, or shUCH-L1#2 cells were loaded in each well of 8-16% SDS-PAGE gel (Invitrogen™ XP08165BOX) and transferred onto a 0.22 um nitrocellulose membrane (GVS Life Sciences, 1212632). Membrane was blocked with 5% non-fat milk at room temperature for 1 hour and incubated with primary antibody (Santa Cruz Biotechnology-anti-UCH-L1 (1:1000) sc-271639 or anti-GAPDH (1:1000) sc-47724) overnight at 4°C. Then the membrane was washed three times with PBS-T (0.1% Tween 20) and incubated with HRP conjugated mouse secondary antibodies (1:5000, Fisher Scientific, PI31432) and developed with ECL Western Blotting Substrate (Thermo Fisher Scientific).

Immunohistochemistry

FFPE tissue sections were incubated at 65°C for 60 minutes and rehydrated in degrading percentage of ethanol. For H&E staining, slides were stained with hematoxylin for 1 minute and eosin for one minute. For immunohistochemical staining, tissue slides were steamed in 10 mM citrate buffer (pH = 6.0) at 95°C for 30 minutes. Slices were then incubated with 3% hydrogen peroxide for 5 minutes. After three times washing with water, tissue sections were blocked with 2.5% goat serum at room temperature for 1 hour. Slices were incubated with primary antibodies (anti-UCH-L1, sc-271639, 1:100; anti-Ki67, sc-23900, 1:100, Santa Cruz Biotechnology) at 4°C overnight. After three times of washing with PBS, slides were incubated with ImmPRESS^® HRP secondary Antibody (MP-7452-50, Vector Laboratories) at room temperature for one hour and developed with DAB kit (Dako).

Animal models

All animal experimental procedures were approved by University of California, Los Angeles, on Laboratory Animal Care (APLAC), IACUC and all animal experiments were conducted in accordance with the Animal Ethics Committee at University of California, Los Angeles. 6-8-week-old male NSG (NOD-SCID-IL2Rγ-null) mice (Jackson Laboratory) were used in this study. Mice were housed at 65-75°F with 40-60% humidity at animal facility with veterinary care provided by the Division of Laboratory Animal Medicine at University of California, Los Angeles.

Subcutaneous xenograft model

DU-145-shCtl, shUCH-L1#1, or shUCH-L1#2 (1×10⁶) cells were mixed with 80% of Matrigel in medium. Cells were subcutaneously implanted into dorsal flanks of male NSG mice. Tumor volumes were measured every six days and calculated by the equation (length × width × height)/2. At day 42 after cell implantation, subcutaneous tumors were harvested and weighed at the end point. Tumors were fixed in 10% formalin and processed and embedded in paraffin or histological analysis.

Spontaneous metastasis model

Male NSG mice were injected subcutaneously with DU-145-RFP-Luc cells expressing either shUCH-L1 or control shRNA (shCtl) on the right dorse flank. Tumor volumes were measured every six days. When tumor volumes of shCtl group reached 400 mm³, tumors from both groups were removed by surgery. Mice were monitored by bioluminescence imaging after surgery, followed by once-a-week post-surgery for three weeks. After three weeks, organs including lungs, liver, kidneys, bones, and lymph nodes were harvested from mice after euthanizing. Organs were fixed in 10% formalin and stored in 70% ethanol for fluorescence imaging. Organs were imaged by fluorescence microscope (Stereo microscopes, Leica, Cat#M205) under red fluorescence channel and brightfield channel. The metastasis in diverse organs were quantified by red fluorescence signals.

Intracardiac injection model

For UCH-L1 knockdown cell lines, DU-145-RFP-Luc-shCtl or DU-145-RFP-Luc-shUCH-L1 cells (1×10⁵ per mouse) were injected into the left ventricle of the heart of 8-weeks-old NSG male mice. Diverse organs including lungs, liver, kidneys, and bones were harvested at Day 21 post-injection. For drug treatment, DU-145-RFP-Luc (1×10⁵ per mouse) were injected into left ventricle of the heart of 8-weeks-old-NSG male mice. Mice were randomized into two groups based on bioluminescence signals at day 3 post injection into two groups and treated with vehicle or LDN-57444 (5 mg/kg, corn oil, i.p., daily) for 14 days. Organs including lungs, liver, kidneys, and bones were harvested at Day 14. All organs were fixed in 10% formalin overnight and stored in 70% ethanol for histopathologic assessment. Metastatic nodules were quantified by quantitative fluorescence images.

LC-MS/MS analysis

Proteomics analysis was on 25 µg of extracted protein from each sample - flash-frozen DU-145 shControl and shUCHL1 xenografts (n = 2 per control condition, and n = 4 per shUCHL1 condition). The flash-frozen xenograft tissues were homogenized in 500 µL lysis buffer consisting of 2% sodium dodecyl sulfate (Fisher Scientific) and 1X protease inhibitor (Sigma Aldrich) using a PRO-250 (ProScientific) homogenizer probe followed by sonication. The insoluble fraction was pelleted by centrifuging the xenograft lysates at 14,000 g for 10 minutes at 4°C. Extracted protein was quantified using a bicinchoninic acid (BCA) protein assay (Thermo Fisher Scientific). Proteins were reduced with a final concentration of 10 mM Tris(2-carboxyethyl) phosphine (Sigma Aldrich) and incubated at room temperature for 1 hour. The free thiol groups were alkylated by adding iodoacetamide (Acros Organics) at a 1.5-fold molar excess of Tris(2-carboxyethyl) phosphine and incubated for 45 minutes at room temperature in the dark. Samples were digested at 37°C overnight with 1 µg sequencing grade modified trypsin enzyme (Thermo Fisher Scientific). Subsequently tryptic peptides were dried on a speed vacuum and reconstituted in 50 µl of 0.1% formic acid (Fisher Scientific) in HPLC grade water (Fisher Scientific) for LC/MS analysis. Tryptic peptides (2 µg) were loaded onto an Acclaim PepMap C18 trap column (Thermo Fisher Scientific) coupled to a Dionex Ultimate Rapid Separation Liquid Chromatography system (Thermo Fisher Scientific) at a rate of 5 µl/min for 10 minutes. A reverse-phased liquid chromatography gradient was used to separate the tryptic peptide on a 25 cm C18 analytical column (New Objective) packed in-house with Magic C18 AQ resin (Michrom Bioresources). The chromatographic program consisted on setting the flow rate to 0.5 µL/min throughout the gradient and changing the mixture of mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in acetonitrile) as follows: Mobile phase B was held at 2% B for 10 minutes with a gradual increase to 35% B for the next 110 minutes followed by a rapid increase to 85% B over 5 minutes, and ending with a 10-minute hold with 2% mobile phase B for column equilibration. Eluting peptides were directly ionized with 2.0 kV on a nanospray Flex Ion ESI source coupled to a LTQ-Orbitrap Elite mass spectrometer (Thermo Fisher Scientific). Each biological sample was analyzed in triplicate. The top ten most abundant ions per MS1 scan were selected for higher energy collision induced dissociation (CID) with a collision energy set to 35 eV and mass resolution to 60,000. The FT AGC target was set to 1e6 and the scan range set to 400-1800 m/z. The MS2 AGC target was 3e4 and dynamic exclusion was enabled for 30 seconds.

Proteomic statistical analysis

For each LC-MS run, the resulting raw data files were searched using Byonic 2.11.0 (Protein Metrics) Swiss-Prot databases. First, containing reference human proteome (2020; 20,626 entries), and again using reference mouse proteome (2020; 17,282 entries). Parameters included trypsin digestion with a maximum of two missed cleavages and precursor mass tolerance of 10 ppm, and 0.5 Da for fragment masses. Fixed cysteine carbamidomethylation and variable methionine oxidation and asparagine deamination were also specified. Peptide identification was filtered to remove hits with > 1% false discovery rate (FDR). Peptides that overlapped in human and mouse searches were removed for a conservative analysis of only human-specific proteins using an in-house R script, for each of the two biological replicates, and three injections per sample. Quantitative values were extracted from MS1 spectra from all identified peptides using an in-house R script based on MSnbase package [85], and abundance changes analyzed using Generic Integration Algorithm. Calculation of statistical weight was performed at spectrum level using WSPP model [86]. Final statistical analysis was performed using student’s T-test, considering only proteins having a p-value less than 1%. These proteins were analyzed on String network analysis (String-db.org), and Enrichr (https://maayanlab.cloud/Enrichr/).

Statistics

Student’s t-test was performed to compare two groups unless otherwise noted. For all, p-values of 0.05 or less were considered statistically significant. **** = P < 0.0001, *** = P < 0.005, ** = P < 0.01, * = P < 0.05, and n.s. = not significant.

Disclosure of conflict of interest

None.