Nuclear localization of proteasome subunit beta 5 serves as an independent prognostic biomarker and promotes invasion in colorectal cancer

Abstract

BACKGROUND

The proteasome subunit beta (PSMB) 5 is a core catalytic subunit of the 20S proteasome. While its overexpression is noted in some cancers, the clinical significance of its subcellular localization in colorectal cancer (CRC) remains unexplored.

AIM

To investigate the subcellular localization of PSMB5, its clinical prognostic value, and functional role in invasion and migration in CRC.

METHODS

Immunohistochemistry, reverse transcription quantitative real-time polymerase chain reaction, western blot, and subcellular fractionation were used to investigate the PSMB5 expression and localization in CRC tissues (n = 129) and cell lines. The prognostic value was assessed using Kaplan-Meier and Cox regression analyses. The PSMB5 functional role was examined using small interfering RNA-mediated knockdown followed by invasion and migration assays.

RESULTS

PSMB5 was significantly overexpressed in tumor tissue clinical samples compared to adjacent normal mucosa. Additionally, it showed heterogeneous subcellular localization (cytoplasmic, nuclear, or both). The high expression and nuclear localization were strongly associated with aggressive clinicopathological features that included poor differentiation, deep invasion, metastasis, and an advanced tumor, node, metastasis stage. Moreover, nuclear PSMB5 localization was identified using a multivariate analysis as an independent prognostic factor for poor overall survival. In vitro, PSMB5 was overexpressed at both the mRNA and protein levels in CRC cell lines; and its nuclear enrichment was markedly heightened in cell lines with metastatic potential. Functionally, the silencing of PSMB5 robustly inhibited the invasive and migratory capacities of CRC cells.

CONCLUSION

Nuclear-localized PSMB5, beyond mere overexpression, may serve as a novel independent prognostic biomarker, a functional requirement for invasion/migration, and a potential therapeutic target in CRC.

Key Words: Colorectal cancer; Proteasome subunit beta 5; Nuclear localization; Prognostic biomarker; Cell migration and invasion; Subcellular fractionation; Small interfering RNA knockdown

Core Tip: This is the first study to investigate the subcellular localization of proteasome subunit beta (PSMB) 5 in colorectal cancer (CRC). We demonstrate that nuclear PSMB5 localization, rather than total PSMB5 expression, serves as an independent prognostic biomarker for poor overall survival in CRC patients. Functional assays show that PSMB5 silencing significantly impairs CRC cell invasion and migration. Notably, metastatic CRC cell lines exhibit higher nuclear-to-cytoplasmic PSMB5 ratios than non-metastatic counterparts. These findings establish nuclear-localized PSMB5 as a novel prognostic indicator and a potential therapeutic target in CRC.

INTRODUCTION

Colorectal cancer (CRC) remains a major global health issue, with increasing incidence and mortality rates, particularly in regions with high socioeconomic development[1]. According to the latest global cancer statistics, CRC accounts for approximately 1.9 million new cases and 0.9 million deaths annually worldwide. This makes it the third most commonly diagnosed malignancy and the second leading cause of cancer-related death[1]. In China, it ranks second and fourth in incidence and mortality, respectively[2]. Progress has occurred in the integration of multidisciplinary approaches that include surgery, chemotherapy, targeted therapy, immunotherapy, and personalized medicine[3]; however, resistance to chemotherapy and targeted therapies remains a significant challenge. Therefore, there is an urgent need for reliable and efficient biomarkers in clinical practice.

The ubiquitin-proteasome system (UPS) is a multi-subunit proteolytic complex responsible for ubiquitin-dependent protein degradation in eukaryotic cells[4]. It regulates key cellular processes such as DNA repair, cell proliferation, survival, and differentiation, and drug resistance. It plays a vital role in the maintenance of cellular homeostasis[5,6]. UPS dysregulation can contribute to various diseases, including cancer. The 26S proteasome is the most prevalent form in eukaryotes and it consists of two 19S regulatory particles and a barrel-shaped catalytic core particle, also known as the “20S proteasome”. This complex is composed of four stacked rings, each made of seven subunits, that can be either α (proteasome subunit alpha 1-7) or β-type (PSMB1-7) that share an α/β sandwich-fold structure[7]. An elevated mRNA expression of PSMB subunits has been observed in neoplastic tissues compared to normal tissues[8,9]. Additionally, PSMB8 downregulation has been shown to inhibit proliferation, migration, and invasion of clear cell renal carcinoma cells[10]. PSMB2 has also been shown to play an oncogenic role in glioma and correlate with the immune microenvironment[11]. PSMB2 knockdown is known to suppress proteasome activity, inhibit cell proliferation, promote apoptosis, and block nuclear respiratory factor 1 activation in gastric cancer cells[12]. PSMB4 overexpression is known to enhance breast cancer cell growth and viability, leading to a poor prognosis[13]. These findings collectively underscore the PSMB family’s significant, yet heterogeneous, contributions to tumorigenesis across cancer types.

As a core subunit of the 20S proteasome, PSMB5 is essential for the rate-limiting step during the hydrolysis of ubiquitin-conjugated proteins due to its chymotrypsin-like activity[14,15]. It plays a crucial role in the elimination of misfolded proteins and intracellular protein level regulation[16]. Recent studies have indicated that PSMB5 is upregulated in several cancers[17-19], including hepatocellular, prostate, and breast cancers, and its dysregulation is associated with tumor cell proliferation, metastasis, immunosuppression, and bortezomib (BTZ) resistance[17-20]. Notably, emerging evidence across different cancer types has begun to establish PSMB5 as a promising prognostic marker. For instance, Plakoula et al[21] (2025) demonstrated that lower PSMB5 levels in bone marrow mononuclear cells were associated with prolonged disease-free survival in patients with BTZ-resistant multiple myeloma . Concurrently, an integrative multi-omics study identified PSMB5 as a key prognostic gene in bladder cancer, showing that its knockdown inhibited tumor cell proliferation and migration[22]. Notably, while the oncogenic functions of PSMB5 are emerging, its role in CRC remains largely unexplored. Moreover, the clinical and biological implications of its subcellular localization pattern have not been assessed.

In this study, we aim to comprehensively investigate the clinicopathological and prognostic significance of PSMB5 in CRC, with a particular focus on its subcellular localization, which was a previously overlooked aspect. To accomplish this, we first examined its expression and subcellular distribution in clinical specimens to evaluate their correlation with clinicopathological features and patient prognosis. In addition, we examined these factors in a panel of CRC cell lines to establish an in vitro model. We then employed a loss-of-function approach to validate the PSMB5 functional role in driving CRC cell invasion and migration. Finally, by integrating our results, we synthesized a testable mechanistic model that linked PSMB5 nuclear accumulation to aggressive CRC progression.

MATERIALS AND METHODS

Patients and specimens

A total of 129 paraffin-embedded CRC tissue samples were retrospectively collected between 2016 and 2018 from the Department of Pathology at Jiangxi Provincial People’s Hospital that is affiliated with the First Affiliated Hospital of Nanchang Medical College (Nanchang, China). The patients were followed-up with for 5 years. This study adhered to ethical standards and was approved by the Ethics Committee of Jiangxi Provincial People’s Hospital [Approval No. KeKuai-2024-(15)], with all participants providing informed consent, which granted a waiver of informed consent due to the retrospective nature of the study and the use of de-identified archived tissue specimens.

Cell lines

Seven CRC cell lines (LoVo, SW620, SW480, HCT116, HT29, RKO, and LS174T) and one immortalized normal colorectal epithelial cell line (FHC) were obtained from the Cell Bank of the Chinese Academy of Sciences, Shanghai, China. SW480 and HCT116 were cultured in the Roswell Park Memorial Institute Medium-1640 medium (Hyclone, UT, United States) with 10% fetal bovine serum (FBS; Gibco, NY, United States), whereas SW620, RKO, and LS174T were cultured in the Dulbecco’s Modified Eagle Medium (Hyclone, UT, United States) with 10% FBS. The LoVo and HT29 cells were cultured in F12K and McCOY’s 5A media, respectively, also with 10% FBS. The FHC was maintained in a Dulbecco’s Modified Eagle Medium/F12 medium with 10% FBS. Among these, LoVo and SW620 are derived from metastatic sites, while SW480, HCT116, HT29, RKO, and LS174T were derived from primary tumors. All cell lines were incubated at 37 °C in a humidified atmosphere with 5% CO₂.

Immunohistochemistry and evaluation of PSMB5 expression

Immunohistochemistry (IHC) was performed on 129 paraffin-embedded CRC tissue samples. Briefly, 4-μm-thick sections were deparaffinized, rehydrated, and subjected to antigen retrieval in pH 9.0 Tris-ethylenediaminetetraacetic acid buffer using a pressure cooker. The endogenous peroxidase activity was quenched with 3% H₂O₂. The sections were then incubated with an anti-PSMB5 polyclonal primary antibody (1:300; Abcepta Ltd., Jiangsu Province, China) for 1 hour at room temperature. This was followed by incubation with a horseradish peroxidase-conjugated secondary antibody (Maixin Biotechnology, Fujian Province, China) for 30 minutes at 37 °C. Signal development was performed using diaminobenzidine with hematoxylin counterstaining.

The PSMB5 expression was independently evaluated by three pathologists blinded to clinical data. A semi-quantitative scoring system was employed that integrated the staining intensity and extent according to previously described criteria[23,24]. The intensity was scored as 0 (negative), 1 (weak), 2 (moderate), or 3 (strong). The percentage of positive tumor cells (extent) was scored as 0 (0%), 1 (1%-25%), 2 (26%-50%), 3 (51%-75%), or 4 (76%-100%). The final score (range 0-12) was obtained by multiplying the intensity and extent scores. Based on this, cases were categorized as negative (0-2), weak (3-4), medium (5-8), or strong (9-12). PSMB5 low expression was defined as (-) or (+), i.e., a final score ≤ 4; and PSMB5 high expression was defined as (++) or (+++), i.e., a final score ≥ 5. For the subcellular (nuclear vs cytoplasmic) localization analysis, cases with completely negative PSMB5 expressions (final score = 0) were excluded, as the absence of staining precluded a reliable assessment of the subcellular distribution.

RNA extraction, reverse transcription, and quantitative real-time polymerase chain reaction

The total RNA was extracted from cells or tissues using TRIzol reagent (Invitrogen, CA, United States) following the manufacturer’s protocol. The RNA concentration and purity were assessed using a Nanodrop spectrophotometer (Thermo Fisher Scientific, NY, United States), and the integrity was verified using agarose gel electrophoresis. For cDNA synthesis, 1 μg of the total RNA was reverse transcribed using the PrimeScript reverse transcription reagent Kit with gDNA Eraser (Takara Bio, Shiga, Japan). Quantitative real-time polymerase chain reaction was subsequently performed using the TB Green Premix Ex Taq II kit (Takara Bio, Shiga, Japan) on a StepOnePlus Real-Time Polymerase Chain Reaction System (Applied Biosystems, CA, United States). Each 20-μL reaction contained 10 μL of premix, 0.8 μL of each primer (10 μmol/L, sequences in Table 1), and 2 μL of cDNA. The thermal cycling protocol consisted of an initial denaturation at 95 °C for 30 seconds, followed by 40 cycles of 95 °C for 5 seconds and 60 °C for 30 seconds. A melting curve analysis was conducted to ensure amplification specificity. All reactions were run in technical triplicate with no-template controls. The relative PSMB5 mRNA expression was calculated using the 2^-ΔΔCt method[25], with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the endogenous control.

Table 1 Sequences of primers used for quantitative real-time polymerase chain reaction.

Gene name	Primer sequence (5’ → 3’)	Product size (bp)	GenBank accession No.
PSMB5	Forward: AGC TGC TGG AGG AGT TCA TC	152	NM_002797.4
	Reverse: GGT CCA GGT CCA GAA GTT GT
GAPDH	Forward: GGA GCG AGA TCC CTC CAA AAT	197	NM_002046.7
	Reverse: GGC TGT TGT CAT ACT TCT CAT GG

PSMB5: Proteasome subunit beta 5; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase.

Protein extraction and western blot analysis

Whole-cell protein extraction and western blot: To analyze of total protein expression, cells were washed twice with ice-cold PBS and lysed in a radioimmunoprecipitation assaybuffer supplemented with a 0.1% protease inhibitor cocktail and 0.5% phenylmethylsulfonyl fluoride (Keygentec, Jiangsu Province, China). Protein concentrations were quantified using the a bicinchoninic acid assay. Equal amounts of protein (50 μg per lane) were separated using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes (Merck Millipore, Burlington, MA, United States). After blocking with 5% non-fat milk in Tris-buffered saline with Tween-20 for 1 hour at room temperature, the membranes were incubated overnight at 4 °C with the following primary antibodies: Anti-PSMB5 (1:300; Abcepta Ltd., CA, United States) and GAPDH (1:500; Abbkine Scientific Co., Ltd., GA, United States). Following three washes with Tris-buffered saline with Tween-20, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (1:2000; Abbkine Scientific Co., Ltd., GA, United States) for 1 hour at room temperature. Protein bands were visualized using BeyoECL Plus chemiluminescent detection reagents (Beyotime Biotechnology, Shanghai, China) according to the manufacturer’s instructions.

Subcellular fractionation: To obtain the cytoplasmic and nuclear protein fractions, the Nuclear and Cytoplasmic Protein Extraction Kit (catalog # P0027; Beyotime Biotechnology, Shanghai, China) was used according to the manufacturer’s instructions. Briefly, approximately 1 × 10⁶ cells were harvested, washed with ice-cold PBS, and resuspended in 200 μL of Cytoplasmic Extraction Reagent A containing 1 mmol/L phenylmethylsulfonyl fluoride. After vigorous vortexing and incubation on ice for 15 minutes, 10 μL of cytoplasmic extraction reagent B was added. The supernatant, containing the cytoplasmic fraction, was collected following centrifugation at 12000 × g for 5 minutes at 4 °C. The remaining pellet was then resuspended in 50 μL of nuclear extraction reagent, vortexed thoroughly, and incubated on ice for 30 minutes with intermittent vortexing. The nuclear fraction was obtained by centrifugation at 12000 × g for 10 minutes at 4 °C. The protein concentration of each fraction was determined using the bicinchoninic acid assay. A western blot analysis was then performed as described in previous subsection, with anti-Lamin B1 (1:1000; Proteintech Group, Inc., IL, United States) used as an additional nuclear loading control.

RNA interference and transfection

Loss-of-function studies were the performed to investigate the PSMB5 functional role. To rigorously minimize potential off-target effects, a pool of three distinct small interfering RNAs (siRNAs) specifically targeting human PSMB5 (small interfering PSMB5-pool) was used as the primary knockdown approach (sequences listed in Table 2). A non-targeting control siRNA [designated as “Scramble”] with the sequence 5’-UUCUCCGAACGUGUCACGUdTdT-3’ (where ‘dT’ denotes deoxythymidine) was purchased from GenePharma (Shanghai, China). The siRNA sequences within the pool are proprietary (GenePharma, Shanghai, China) and designed for maximal knockdown efficiency and specificity. For transfection, the CRC cells were seeded and allowed to adhere overnight. Complexes were formed by mixing the siRNA (final concentration 20 nM for the pool or Scramble) with Lipofectamine RNAiMAX transfection reagent (Invitrogen, CA, United States) in an Opti-MEM and then added to the cells. Cells were harvested for RNA and protein extraction 48 hours post-transfection to verify the knockdown efficiency, and for functional assays 72 hours post-transfection.

Table 2 Sequences of the small interfering RNAs in the proteasome subunit beta 5-targeting pool.

siRNA name/ID	Target sequence (DNA notation)	Actual synthesized oligo (5’ → 3’)	Target position
siPSMB5-#1	TCGAATCTATGAGCTTCGAAATA	UCGAAUCUAUGAGCUUCGAAAUAdTdT	385-407
siPSMB5-#2	GTGCTTGAAACCTAAGTCATTTG	GUGCUUGAAACCUAAGUCAUUUGdTdT	538-560
siPSMB5-#3	GACTGTCATTGGTAATACGGACA	GACUGUCAUUGGUAAUACGGACAdTdT	935-955
Scramble (siControl)	TTCTCCGAACGTGTCACGTTT	UUCUCCGAACGUGUCACGUdTdT	N/A

All small interfering RNAs were synthesized as RNA oligos with 3’dTdT overhangs. si: Small interfering; PSMB5: Proteasome subunit beta 5; N/A: Not applicable.

The initial confirmatory experiments using two independent short hairpin RNAs yielded consistent results. The sequences and detailed experimental procedures are provided in Supplementary Table 1 and the Supplementary material, respectively.

Cell invasion and migration assays

We did not select LoVo for the loss-of-function experiments due to its extremely low transfection efficiency (< 15%), despite repeated optimization attempts. Instead, we selected the isogenic pair SW480 (primary CRC tumor-derived) and SW620 (lymph node metastasis-derived from the same patient), which share a similar genetic background and, more importantly, exhibit high transfection efficiency - ensuring robust and reproducible functional data for subsequent invasion and migration assays.

For the Transwell invasion assay, 72 hours post-transfection, 1 × 10⁵ cells in a serum-free medium were seeded into the upper chambers of Matrigel-coated inserts (Corning, NY, United States). A complete medium containing 10% FBS was added to the lower chamber as a chemoattractant. After 24 hours of incubation, cells that invaded through the Matrigel were fixed, stained with 0.1% crystal violet, and counted under a microscope in five random fields.

For the wound healing migration assay, 72 hours post-transfection, a uniform scratch was created in a confluent cell monolayer using a sterile 200-μL pipette tip. The width of the scratch was photographed at 0 hour and 24 hours under a microscope. The migration rate was calculated as the percentage of wound closure.

All functional assays were performed in triplicate and repeated at least three independent times using cells transfected with the small interfering PSMB5-pool or Scramble.

Statistical analysis

The statistical analyses were performed using the SPSS version 27.0 (Statistical Package for the Social Sciences; IBM Corp., Armonk, NY, United States). Normality of continuous data was assessed using the Shapiro-Wilk test. The Mann-Whitney U test was used to compare the PSMB5 protein expression between the CRC tissues and matched adjacent mucosa tissues. It was also used to compare the expressions among CRC tissues, lymph node metastatic tissues, and distant metastatic tissues. The χ² test was used to analyze the relationship between PSMB5 expression and clinicopathological parameters. A Kaplan-Meier survival curve analysis was used to evaluate the association between PSMB5 expression and patient prognosis. The Cox proportional hazards regression model was used to identify factors that influenced the CRC prognosis. PSMB5 expression in CRC cell lines was analyzed using a one-way analysis of variance followed by a Tukey’s post hoc test. Differences between the two groups in functional assays were assessed using a Student’s t-test. A two-sided P < 0.05 was considered statistically significant.

RESULTS

Elevated expression and nuclear localization of PSMB5 correlate with malignant progression and poor prognosis in CRC

To investigate PSMB5 expression in CRC and its clinicopathological relevance, we performed an IHC analysis on a cohort of paraffin-embedded samples that included 129 primary CRC cases with matched adjacent normal mucosa, 56 lymph node metastases, and 4 distant metastases (Table 3). Notably, although 29 distant metastases were clinically documented, only 4 were surgically resected and available for pathological evaluation. The IHC revealed that PSMB5 was localized to both the cytoplasm and nucleus of the CRC cells (Figure 1). Strong immunostaining was observed in primary CRC tissues (Figure 1A and B), whereas the adjacent normal mucosa exhibited minimal or faint staining (Figure 1C). The positive expression rate of PSMB5 in CRC tissues was 97.7%, with strong, moderate, and weak positivity rates of 38.8%, 37.2%, and 21.7%, respectively (Table 3), which was significantly higher than the 30.2% positivity rate in the adjacent normal mucosa (P < 0.01). Moreover, PSMB5 expression was significantly elevated in metastatic lymph node carcinomas (Figure 1D and Table 3, P = 0.04) and showed a non-significant increasing trend in liver metastases (Table 3, P > 0.05) compared to primary tumors (Figure 1A). Among the lymph node metastases, 64.3% exhibited strong positivity (Table 3). All four liver metastasis cases exhibited moderate to strong PSMB5 expression (Figure 1E and F). While the comparison did not reach statistical significance (n = 4, P > 0.05), this small-sample observation suggests that further study in larger cohorts is warranted to examine whether an elevated PSMB5 expression is associated with metastatic progression in CRC. A χ² analysis demonstrated that high PSMB5 expression was significantly associated with poor tumor differentiation (P = 0.002), deeper invasion depth (P = 0.006), lymph node metastasis (P = 0.001), distant metastasis (P = 0.001), and an advanced tumor, node, and metastasis (TNM) stage (P = 0.001). No significant correlations were found with age (P = 0.458), gender (P = 0.954), or tumor size (P = 0.798; Table 4).

Figure 1 Immunohistochemical detection of proteasome subunit beta 5 protein expression in colorectal cancer tissues. A: Well to moderately differentiated primary carcinoma; B: Poorly differentiated primary carcinoma; C: Peritumoral mucosa tissue; D: Lymph node metastasis; E and F: Liver metastasis (two representative cases). Scale bars: 200 μm (4 ×) and 50 μm (20 ×).

Table 3 Proteasome subunit beta 5 expression in primary colorectal cancer, adjacent normal mucosa, and metastatic tissues.

Tissues	Total	-	+	++	+++	P value
Adjacent normal mucosa	129	90	14	18	7
Primary CRC	129	3	28	48	50	< 0.01b
Lymph node metastasis	56	1	2	17	36	0.04a
Liver metastasis	4	0	0	2	2	> 0.051

^aP < 0.05 lymph node metastasis vs primary colorectal cancer.

^bP < 0.01 primary colorectal cancer vs adjacent normal mucosa.

¹Liver metastasis vs primary colorectal cancer.

CRC: Colorectal cancer.

Table 4 Proteasome subunit beta 5 expression and clinicopathological features in colorectal cancer.

Parameters	Total number (n = 129)	High expression (n = 98)	Low expression (n = 31)	χ2 value	P value
Age (years)
< 60	55	40	15	0.552	0.458
≥ 60	74	58	16
Gender
Male	66	50	16	0.003	0.954
Female	63	48	15
Tumor size (cm)
< 5	64	48	16	0.065	0.798
≥ 5	65	50	15
Differentiation
Well/moderate	105	74	31	9.327	0.002a
Poor	24	24	0
Depth of invasion
T1-T2	21	11	10	7.645	0.006a
T3-T4	108	87	21
Lymph node metastasis
No	73	45	28	18.902	0.001a
Yes	56	53	3
Distant metastasis
No	100	69	31	11.834	0.001a
Yes	29	29	0
TNM stage
I-II	61	33	28	30.320	0.001a
III-IV	68	65	3

^aP < 0.01.

TNM: Tumor, node, and metastasis.

We then examined the subcellular localization pattern of PSMB5. In poorly differentiated CRC tissues, 62.5% of the cases exhibited predominantly nuclear PSMB5 localization, 29.2% showed combined nuclear and cytoplasmic localization, and only 8.3% showed cytoplasmic localization alone (Table 5). In contrast, in well-differentiated and moderately differentiated tissues, PSMB5 was primarily localized to the cytoplasm (55.9%), with 26.5% showing nuclear localization and 17.7% showing both nuclear and cytoplasmic localization. Nuclear localization of PSMB5 was significantly associated with poor differentiation (P = 0.001), deeper invasion (P = 0.002), lymph node metastasis (P = 0.001), distant metastasis (P = 0.001), and advanced TNM stage (P = 0.001), but not with age, gender, or tumor size (all P > 0.05). These data suggest that the nuclear accumulation of PSMB5 is closely linked to the malignant progression of CRC.

Table 5 Proteasome subunit beta 5 localization and clinicopathological features in colorectal cancer.

Parameters	Total number (n = 126)	Cytoplasm (n = 56)	Nucleus (n = 41)	Cytoplasm and nucleus (n = 29)	χ2 value	P value
Age (years)
< 60	54	25	19	10	0.186	0.911
≥ 60	72	34	23	15
Gender
Male	65	30	18	17	2.233	0.442
Female	61	26	23	12
Tumor size (cm)
< 5	62	31	18	13	1.018	0.601
≥ 5	64	28	24	12
Differentiation
Well/moderate	102	57	27	18	18.246	0.001a
Poor	24	2	15	7
Depth of invasion
T1-T2	19	16	2	1	12.566	0.002a
T3-T4	107	43	40	24
Lymph node metastasis
No	71	52	11	8	45.793	0.001a
Yes	55	7	31	17
Distant metastasis
No	97	55	25	17	17.143	0.001a
Yes	29	4	17	8
TNM stage
I-II	59	49	4	6	59.797	0.001a
III-IV	67	10	38	19

^aP < 0.01.

Three cases with completely negative proteasome subunit beta 5 expression by immunohistochemistry were excluded from subcellular localization analysis. TNM: Tumor, node, and metastasis.

The prognostic significance of PSMB5 was evaluated using a Kaplan-Meier survival analysis. Among the 129 CRC cases, 98 exhibited high PSMB5 expression (++ to +++) and 31 showed low expression (- to +; Table 4). Patients with high PSMB5 expression had a significantly lower five-year overall survival rate (67.3%) compared to those with low expression (93.5%; P = 0.003, log-rank test; Figure 2A). This result indicated that an elevated PSMB5 expression correlates with poor prognosis. We further assessed the impact of PSMB5 subcellular localization on patient survival. Patients with nuclear PSMB5 expression had a five-year survival rate of 50.0%, and those with both nuclear and cytoplasmic expression had a five-year survival rate of 56.0%. In striking contrast, patients with only cytoplasmic PSMB5 localization exhibited a markedly higher survival rate of 98.3%. Survival outcomes were significantly worse for patients with nuclear or combined localization than for those with cytoplasmic localization alone (P < 0.001, log-rank test; Figure 2B), underscoring the adverse prognostic value of PSMB5 nuclear localization.

Figure 2 Kaplan-Meier survival analysis of proteasome subunit beta 5 expression level, subcellular localization, and prognosis in colorectal cancer patients. A: Overall survival of colorectal cancer patients stratified by proteasome subunit beta 5 expression levels; B: Overall survival of colorectal cancer patients stratified by proteasome subunit beta 5 subcellular localization patterns. ^aP < 0.01 and ^bP < 0.001, for the comparison in A and for comparisons of the cytoplasm-only group vs the other two groups in B, respectively. PSMB5: Proteasome subunit beta 5.

A univariate Cox regression analysis identified both a high PSMB5 expression and nuclear localization as significant prognostic factors (P = 0.01 and P = 0.001, respectively; Table 6). A multivariate analysis further confirmed that the nuclear localization of PSMB5, together with other pathological parameters, can serve as an independent prognostic factor of overall survival in CRC patients (P = 0.016; Table 6).

Table 6 Univariate and multivariate Cox regression analyses for overall survival in colorectal cancer patients.

Parameters	Univariate Cox regression			Multivariate Cox regression
	HR	95%CI for HR	P value	HR	95%CI for HR	P value
Age	1.716	0.836-3.525	0.141
Gender	0.829	0.422-1.628	0.586
Tumor size	0.677	0.342-1.340	0.263
Differentiation	0.266	0.132-0.532	0.001b	0.548	0.261-1.151	0.112
Depth of invasion	27.467	0.637-1184.629	0.085
Lymph node metastasis	3.599	1.717-7.543	0.001b	0.325	0.120-0.876	0.026a
Distant metastasis	7.417	3.692-14.898	0.001b	2.521	1.071-5.934	0.034a
TNM stage	19.904	4.760-83.235	0.001b	9.455	1.434-62.343	0.020a
PSMB5 expression	6.477	1.549-27.075	0.010a	0.856	0.097-7.515	0.888
PSMB5 localization1			0.001b			0.016a
PSMB5 localization (1)2	41.534	5.576-309.408	0.001b	18.069	2.137-152.804	0.008b
PSMB5 localization (2)3	33.024	5.256-320.203	0.001b	24.432	2.754-216.782	0.004b

^aP < 0.05.

^bP < 0.01.

¹Proteasome subunit beta 5 localization represents the overall comparison among the three expression patterns (nuclear, cytoplasmic, and combined nuclear + cytoplasmic).

²Proteasome subunit beta 5 localization (1) represents the comparison between nuclear expression and cytoplasmic expression.

³Proteasome subunit beta 5 localization (2) represents the comparison between combined nuclear + cytoplasmic expression and cytoplasmic expression.

TNM: Tumor, node, and metastasis; PSMB5: Proteasome subunit beta 5; HR: Hazard ratio; CI: Confidence interval.

Collectively, these findings established that an elevated PSMB5 expression and its nuclear localization are hallmarks of aggressive CRC biology and a poor clinical outcome.

High PSMB5 expression and nuclear enrichment are associated with metastatic potential in CRC cell lines

To explore the relationship between PSMB5 expression and metastatic propensity in vitro, we analyzed a panel of CRC cell lines with distinct metastatic potentials. The panel included two metastatic cell lines (SW620 and LoVo), five-primary tumor-derived cell lines (RKO, SW480, LS174T, HCT116, and HT29), and one immortalized normal colorectal epithelial cell line (FHC). First, a quantitative real-time polymerase chain reaction analysis revealed that the PSMB5 mRNA levels were substantially elevated in all CRC cell lines compared to the FHC control (Figure 3A). The metastatic cell lines SW620 and LoVo exhibited the highest upregulation, with PSMB5 expression increased by 38.8-fold and 67.0-fold, respectively. The primary cell lines also showed marked overexpression that ranged from 13.9-fold (LS174T) to 30-fold (RKO; Figure 3A). These findings indicate that PSMB5 is transcriptionally upregulated in CRC cells, with a pronounced enrichment in lines that possess metastatic capacity.

Figure 3 Proteasome subunit beta 5 is overexpressed and nuclear-enriched in metastatic colorectal cancer cell lines. A: Relative proteasome subunit beta 5 (PSMB5) mRNA levels in a panel of colorectal cancer cell lines compared to the normal colonic epithelial cell line FHC, determined by quantitative real-time polymerase chain reaction. Data are mean ± SD (n = 3). ^aP < 0.05 vs FHC, ^bP < 0.01 vs FHC, ^cP < 0.001 vs FHC; B: Western blot analysis of total PSMB5 protein in whole-cell lysates from the indicated cell lines. GAPDH serves as a loading control; C-E: Subcellular distribution of PSMB5 protein. Western blot analysis of PSMB5 in cytoplasmic and nuclear fractions extracted from the indicated cell lines, GAPDH and Lamin B1 serve as compartment-specific loading controls for cytoplasm and nucleus, respectively. Due to the limited number of lanes per gel, samples were analyzed across multiple independent blots (C); quantitative analysis of PSMB5 nuclear enrichment, nuclear-to-cytoplasmic ratio of PSMB5 in each cell line (D); percentage of total PSMB5 protein localized to the nucleus (E). Data are derived from densitometric analysis of three independent subcellular fractionation experiments and are presented as mean ± SD. ^aP < 0.05, ^bP < 0.01, ^cP < 0.001, CRC cell lines (RKO, SW620, SW480, LoVo, HT29, LS174T, HCT116) vs FHC. PSMB5: Proteasome subunit beta 5; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase; C: Cytoplasmic; N: Nuclear.

This upregulation was confirmed at the protein level using a western blot analysis that detected a specific 28 kDa band corresponding to PSMB5 in all tested cell lines, including FHC (Figure 3B). Consistent with the mRNA data, the PSMB5 protein abundance was notably higher in metastatic cell lines compared to most of the primary lines. Collectively, these results demonstrated a strong positive correlation between PSMB5 expression - both transcriptional and translational - and the metastatic potential of CRC cell lines.

We further asked whether the aggressive phenotype was linked to a specific subcellular distribution of PSMB5. Subcellular fractionation and a western blot analysis revealed distinct compartmentalization patterns (Figure 3C-E, arranged from left to right). While PSMB5 was detectable in both the cytoplasm and nucleus across all of the cell lines (Figure 3C), the metastatic lines SW620 and LoVo exhibited pronounced nuclear accumulation. A quantitative analysis demonstrated that the nuclear-to-cytoplasmic ratio of PSMB5 was significantly higher in these metastatic lines compared to those of the non-metastatic primary lines (RKO, SW480, LS174T, HCT116, HT29) or the normal FHC cells (Figure 3D). This nuclear enrichment was also evident when calculating the percentage of the total PSMB5 protein residing in the nucleus (Figure 3E). Notably, this differential localization occurred independently of the total protein overexpression observed in Figure 3B. This result suggested that nuclear trafficking is a specifically regulated process in aggressive cells.

Collectively, these results demonstrated a strong positive correlation between PSMB5 expression, both transcriptional and translational, and the metastatic potential of CRC cell lines. Additionally, the results further identified the nuclear enrichment of PSMB5 as a distinctive feature of metastatic cells.

Silencing PSMB5 with a siRNA pool reduces its expression at both transcriptional and translational levels

Efficient knockdown of PSMB5 mRNA: Transfection with the small interfering PSMB5-pool resulted in an efficient and significant reduction of the PSMB5 mRNA levels in the SW620 and SW480 cells compared to that of the Scramble, as determined by quantitative real-time polymerase chain reaction (Figure 4A).

Figure 4 Efficient knockdown of proteasome subunit beta 5 in colorectal cancer cells. A: PSMB5 mRNA levels. SW620 and SW480 cells were transfected with a PSMB5-targeting small interfering RNA (siRNA) pool or non-targeting control siRNA. PSMB5 mRNA was significantly reduced by PSMB5-targeting siRNA pool in both cell lines. Gene expression was normalized to GAPDH and is presented relative to non-targeting control siRNA (set as 1). Data are mean ± SD from three independent experiments. ^aP < 0.001 vs non-targeting control siRNA; B and C: PSMB5 protein levels. Whole-cell lysates were analyzed by Western blot (B) with quantitative analysis (C). GAPDH served as a loading control. Data are mean ± SD from three independent experiments. ^aP < 0.001 vs Scr; D: Subcellular distribution. Cytoplasmic and nuclear fractions from SW620 and SW480 cells transfected with PSMB5-targeting siRNA pool or non-targeting control siRNA were analyzed by western blot. Purity of fractions was confirmed using GAPDH (cytoplasmic marker) and Lamin B1 (nuclear marker). Scr: Non-targeting control small interfering RNA; sipool: Proteasome subunit beta 5-targeting small interfering RNA pool; PSMB5: Proteasome subunit beta 5; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase; C: Cytoplasmic; N: Nuclear.

Concurrent depletion of PSMB5 protein in cytoplasmic and nuclear compartments: This transcriptional knockdown led to a pronounced decrease in PSMB5 protein abundance. A western blot analysis of the whole-cell lysates confirmed the protein level reduction (Figure 4B). Moreover, subcellular fractionation followed by western blotting demonstrated that the PSMB5 protein depletion occurred concurrently in both the cytoplasmic and nuclear fractions (Figure 4C-E). The purity of the fractions was verified by the exclusive presence of the cytoplasmic marker GAPDH and the nuclear marker Lamin B1.

Silencing PSMB5 with a siRNA pool Impairs CRC cell invasion and migration

The above results demonstrated that PSMB5, particularly its nuclear localization, was associated with malignant progression and poor prognosis in CRC patients and correlated with metastatic potential in CRC cell lines. Furthermore, we confirmed that the siRNA-mediated knockdown effectively reduced the PSMB5 levels in both the cytoplasmic and nuclear compartments. Based on these findings, we then sought to determine whether PSMB5 functionally contributed to CRC cell invasion and migration.

The results of the Matrigel-coated Transwell assays showed that depletion of PSMB5 significantly attenuated the invasive capacity of both cell lines (Figure 5). Similarly, the wound healing assays revealed that PSMB5 silencing markedly impaired cell migration, as evidenced by a substantially slower wound closure rate (Figure 6). Collectively, these data demonstrated that PSMB5 is a potent promoter of CRC cell invasion and migration in vitro. These key findings were independently corroborated using two distinct short hairpin RNA constructs (Supplementary Figures 1-3).

Figure 5 Effect of proteasome subunit beta 5 knockdown on the invasion of colorectal cancer cells assessed by Transwell assay. A: Representative images of invading SW620 and SW480 cells, respectively, after proteasome subunit beta 5 knockdown; B: Quantitative analysis of invasion for SW480 and SW620 cells transfected with a proteasome subunit beta 5 small interfering RNA pool or non-targeting control mall interfering RNA. Data are presented as mean ± SD; ^aP < 0.05, ^bP < 0.01 vs non-targeting control small interfering RNA group. Scr: Non-targeting control small interfering RNA; sipool: Proteasome subunit beta 5-targeting small interfering RNA pool.

Figure 6 Impact of proteasome subunit beta 5 knockdown on colorectal cancer cell migration assessed by scratch assay. A and B: Representative scratch assay images of SW620 and SW480 cells following proteasome subunit beta 5 silencing; C and D: Quantification of wound closure for SW620 and SW480 cells transfected with a proteasome subunit beta 5 small interfering RNA pool vs non-targeting control mall interfering RNA. Data are presented as mean ± SD; ^aP < 0.001 vs non-targeting control small interfering RNA group. Scr: Non-targeting control small interfering RNA; sipool: Proteasome subunit beta 5-targeting small interfering RNA pool.

DISCUSSION

Metastasis is the primary cause of mortality in CRC[26]. However, effective treatments for metastatic disease remain limited. This therapeutic challenge underscores the urgent need to identify novel biomarkers and drivers of metastatic progression to guide improved diagnostic and therapeutic strategies.

The UPS is the principal pathway for targeted protein degradation in eukaryotic cells, responsible for the turnover of over 80% of intracellular proteins and thus essential for maintaining proteostasis[4]. Beyond protein quality control, the UPS precisely regulates the stability of key proteins involved in critical processes such as DNA repair, cell cycle progression, differentiation, and drug resistance, thereby exerting broad influence on cellular homeostasis[5,6]. The 20S catalytic core particle of the proteasome, composed of α- and β-type subunits, executes the proteolytic activity. Among these, PSMB5 is a core β-subunit endowed with chymotrypsin-like activity, which is rate-limiting for the hydrolysis of ubiquitin-tagged proteins[7]. PSMB5 is dysregulated in multiple cancers, and its overexpression correlates with a poor prognosis in hepatocellular carcinoma and testicular germ cell tumors[17,27]. In addition, it contributes to tumor progression and therapy resistance in prostate cancer and multiple myeloma[18,20]. In breast cancer, PSMB5 silencing has been shown to inhibit cell growth and migration and to modulate immune responses, highlighting its multifaceted oncogenic roles[19]. Recent work has demonstrated that lower PSMB5 levels in bone marrow mononuclear cells correlate with prolonged disease-free survival in BTZ-resistant multiple myeloma patients, further supporting the prognostic value of PSMB5 across different hematological and solid malignancies[21]. However, both the general clinical significance and the specific prognostic value of PSMB5 subcellular localization in CRC remain largely unexplored.

We showed that an elevated PSMB5 protein expression was significantly associated with aggressive features that included poor differentiation, deep tumor invasion, lymph node and distant metastasis, and an advanced TNM stage. Consequently, patients with high PSMB5 expression exhibited markedly worse overall survival. These findings are consistent with the oncogenic profile of PSMB5 reported in other cancer types[17-20,27]. Complementary experiments in CRC cell lines demonstrated that both mRNA and protein levels of PSMB5 are higher in metastatic cell lines compared to non-metastatic or normal colonic epithelial cells, further reinforcing the link between PSMB5 and aggressive CRC phenotypes.

Notably, our study uncovered a critical layer of regulation beyond mere expression levels: The subcellular localization of PSMB5. We observed that PSMB5 can be localized in the cytoplasm, nucleus, or both compartments in CRC cells. Strikingly, nuclear PSMB5 localization (either alone or combined with cytoplasmic staining) was strongly correlated with more aggressive disease characteristics-such as poorer differentiation, metastasis, and advanced stage-and conferred a significantly worse prognosis compared to purely cytoplasmic localization. Multivariate Cox analysis further identified the nuclear PSMB5 expression as an independent prognostic factor. These results suggest that the nuclear translocation of PSMB5 is not a passive consequence but may actively contribute to CRC malignancy. This finding aligns with a very recent integrative multi-omics study in bladder cancer, which identified PSMB5 as a key prognostic gene and demonstrated that its knockdown inhibits tumor cell proliferation and migration, underscoring the conserved oncogenic role of PSMB5 across different tumor types[22].

To functionally validate the pro-metastatic role of PSMB5 suggested by our clinical data, we performed the loss-of-function experiments in vitro. Silencing PSMB5 expression significantly impaired the invasive and migratory capacities of CRC cells. Collectively, these functional and clinical findings establish PSMB5 as a promoter of CRC cell motility and invasion. Having established its functional necessity, we sought to understand the cellular basis for its association with metastasis.

Importantly, our complementary analysis in CRC cell lines provided a mechanistic clue that bridges the clinical observation and the functional phenotype. We found that metastatic cell lines (SW620, LoVo) not only overexpress PSMB5 but also display a significantly higher nuclear-to-cytoplasmic ratio compared to their non-metastatic counterparts. This cell-autonomous nuclear accumulation recapitulates the nuclear localization pattern observed in the aggressive primary tumors and lymph node metastases. It strongly suggests that the nuclear translocation of PSMB5 is a regulated process associated with the acquisition of metastatic competence, rather than a passive consequence of generalized protein overexpression.

The mechanistic implications of PSMB5 nuclear localization are intriguing. While its canonical role resides in cytoplasmic protein turnover, the nuclear proteasome is known to perform specialized functions, including the spatially and temporally controlled degradation of nuclear regulatory proteins such as transcription factors[28]. For instance, the epithelial-mesenchymal transition (EMT)-transcription factor snail (SNAI1) is a canonical substrate of the UPS, and its targeted degradation has been shown to suppress EMT in gastrointestinal cancers[29]. Consistently, studies in other cancer types have shown that pharmacological targeting of the proteasome complex suppress cancer progression by modulating the stability of EMT regulators such as snail[30]. In our study, we observed that metastatic CRC cell lines and aggressive patient tumors exhibit nuclear enrichment of PSMB5. Conversely, targeted depletion of PSMB5 via siRNA led to a marked reduction of the protein in both nuclear and cytoplasmic compartments, as confirmed by western blotting. Based on these observations together with the established role of the proteasome in regulating EMT-related transcription factors, we speculate cautiously that the nuclear enrichment of PSMB5 might drive the aberrant degradation of certain transcription factors, which could potentially activate gene expression programs conducive to EMT and metastatic dissemination. However, it is important to emphasize that this mechanistic hypothesis is not directly tested in the present study, and the references cited involve different proteins and different cancer types. Nevertheless, the observed correlation between nuclear PSMB5 localization and poor prognosis, raises the hypothesis that systematic identification of PSMB5’s nuclear substrates represent an important direction for future investigation.

One limitation of this study should be acknowledged. Although our clinical and cellular data consistently supported the role of PSMB5 as a prognostic biomarker in CRC, we did not perform in vivo xenograft experiments to validate its function in tumor growth or metastasis. Future studies using animal models, such as subcutaneous or orthotopic xenografts of PSMB5-knockdown CRC cells in nude mice, are warranted to further corroborate the biological significance of PSMB5 in CRC progression.

Additionally, while the precise nuclear substrates remain to be identified, this spatial redistribution suggests a profound rewiring of nuclear proteostasis in advanced CRC. Therefore, future studies aimed at deciphering the exact molecular mechanisms - for instance, by defining the nuclear interactome of PSMB5 via immunoprecipitation mass spectrometry or tracking its dynamics with live-cell imaging - will be crucial to determine whether it promotes metastasis by degrading tumor suppressors or stabilizing oncogenes.

CONCLUSION

In conclusion, the nuclear localization of PSMB5 - rather than its total expression level - serves as a robust and independent prognostic biomarker in CRC. It is also functionally required for the invasive and migratory capacities of CRC cells. Together, these findings establish nuclear-localized PSMB5 as a promising therapeutic target.

ACKNOWLEDGEMENTS

The authors thank Mr. Feng Xiong from Department of Pathology, and Mr. Chang-Hui Xiong from Department of Science and Education, Jiangxi Provincial People’s Hospital, The First Affiliated Hospital of Nanchang Medical College, for their valuable assistance (technical assistance in preparing histological sections and guidance on statistical analysis, respectively).