UBE2I promotes immune infiltration and tumor progression in thyroid cancer and modulates hnRNPA2B1 SUMOylation

Abstract

Thyroid cancer (THCA), a frequent endocrine tumor, has been on the rise in recent years. Identifying potential targets and further clarifying the underlying molecular mechanisms are essential for improving the management of THCA. Multiple genetic and epigenetic changes influence the onset and course of THCA. SUMOylation is a critical posttranslational modification (PTM) that has been reported to be involved in THCA. Ubiquitin-conjugating enzyme E2I (UBE2I), which is required for SUMO conjugation, is crucial to the development of different malignancies. Nonetheless, its role in THCA is not completely understood. In this study, we found that UBE2I was upregulated in THCA. Knock-in and knockdown results showed that UBE2I promoted cell growth, migration, and invasion, as well as macrophage M2 polarization in THCA cells. Mechanistically, UBE2I enhanced the localization of hnRNPA2B1 in the cytoplasm by promoting its SUMOylation, and the inhibition of SUMOylation reversed the tumor-promoting effect of UBE2I overexpression in THCA cells. In vivo, downregulation of UBE2I reduced tumor growth and the level of CD206. Taken together, our findings suggest that UBE2I promotes growth, migration, invasion, and M2 polarization of macrophages in THCA cells, potentially through a mechanism involving SUMOylation and the subsequent cytoplasmic localization of hnRNPA2B1.

With its origins in the thyroid gland, THCA has the highest prevalence among endocrine cancers (1). The prevalence of THCA has increased during the last few decades (2), and it is predicted that by 2030, THCA would rank ninth in men and second in women among cancers, which would substantially increase the health and financial burden on society (3). In China, THCA has a remarkable yearly growth rate of 20% (4), and over 3.7 million new cases are expected to be identified in mainland China between 2028 and 2032 (5). With advances in THCA diagnosis and therapy (6), most individuals with THCA can be treated through various approaches, such as surgery or 131I; however, a small proportion of patients still succumb from metastasis or recurrence (7). It has been reported that 28% of patients with THCA experience relapses and that 9% of THCA patients die (8). These trends highlight the continuing need to strengthen the diagnosis, prevention, and treatment of THCA (6, 9).

THCA has a complicated etiology, with multiple factors contributing to its development. Multiple genetic and epigenetic changes influence the onset and course of THCA (10). Small ubiquitin-like modifier (SUMO) binding, also known as SUMOylation, is a critical posttranslational modification that regulates intracellular transport and signal transduction by affecting subcellular localization and protein stability (11, 12). This process involves the covalent attachment of SUMO proteins, such as SUMO1, SUMO2, and SUMO3, to specific lysine residues on target proteins, with the participation of SAE1/SAE2 and UBC9 (13). Through these, SUMOylation contributes to many essential biological processes, including cell division, protein degradation, and DNA repair (14, 15). Aberrant regulation of SUMOylation has been linked to multiple human cancers, including THCA (16, 17, 18). Papillary THCA is characterized by altered expression of genes related to the SUMOylation process (19). De Andrade JP et al. (20) reported that anaplastic thyroid cancer may benefit from novel treatment that targets the SUMO pathway. Therefore, correcting dysregulated SUMOylation may contribute to improved management of THCA.

Located at human chromosome band 16p13.3, UBE2I encodes the E2 enzyme that accelerates the SUMOylation process and is also known as UBC9 (21). Previous studies have shown that UBE2I exerts multiple physiopathological roles. For instance, UBE2I is necessary for promoting and preserving stem cell pluripotency (22). In ovarian folliculogenesis, UBE2I deficiency in oocytes leads to infertility in mice (23). In cultured vascular smooth muscle cells, UBE2I affects proliferation and migration (24). In addition, loss of UBE2I in adipocytes causes lipoatrophy in mice (25). Moreover, UBE2I has been implicated in several tumors. In hepatocellular carcinoma, UBE2I has been found to promote metastasis and is associated with poor prognosis (26). UBE2I also contributes to hepatocellular carcinoma cell invasion, migration, and growth through an autophagy-associated pathway, further linking it to poor prognosis (27). In cholangiocarcinoma, UBE2I is upregulated and enhances tumorigenesis and chemoresistance (28). Nevertheless, the function of UBE2I in THCA remains unclear.

Therefore, this work aimed to investigate the function and underlying mechanism of UBE2I in THCA. Loss- and gain-of-function results confirmed that UBE2I promoted proliferation, migration, invasion, and macrophage polarization in THCA cells. Mechanistically, UBE2I enhanced the cytoplasmic localization of hnRNPA2B1 by promoting its SUMOylation, and the inhibition of SUMOylation reversed the tumor-promoting effects of UBE2I overexpression in THCA cells. In vivo, downregulation of UBE2I reduced tumor growth and decreased the level of CD206. Together, these results support a molecular mechanism in which the UBE2I/hnRNPA2B1 regulatory axis enhances SUMOylation-dependent THCA progression and suggest that UBE2I may represent a therapeutically relevant target in THCA.

Results

Highly expressed UBE2I in THCA

To explore the role of UBE2I in THCA progression, UBE2I expression was first examined in THCA. Based on analyses of the UALCAN and GEPIA databases, UBE2I expression was markedly elevated in THCA (Fig. 1, A and B). Similar results were also found in the GSE129562 microarray dataset (Fig. 1C). To further confirm these findings, we collected 43 pairs of tumor tissues and matched adjacent non-tumorous thyroid tissues from patients with THCA. Both protein and mRNA levels of UBE2I were significantly increased in THCA tissues compared with those in para-carcinoma tissues (Fig. 1, D and E). Additionally, UBE2I expression was consistently upregulated in three THCA cell lines, and UBE2I expression in TPC-1 and CAL62 cells was higher than that in IHH-4 cells (Fig. 1F). In addition, Nthy-ori 3-1 cells with relatively low endogenous UBE2I expression were used as a comparison control to support assays performed in TPC-1 and CAL62 cells. Thus, subsequent functional experiments were conducted in TPC-1 and CAL62 cells. Together, these results indicate that UBE2I is highly expressed in THCA.

Figure 1.

UBE2I was highly expressed in THCA.A, the expression profile of UBE2I based on the UALCAN database. ∗∗∗p < 0.001. B, the expression profile of UBE2I based on the GEPIA database. ∗p < 0.05. C, the expression profile of UBE2I based on the GSE129562 microarray dataset. ∗p < 0.05. D, the relative mRNA expression of UBE2I in 43 pairs of tumor tissues and matched adjacent non-tumorous thyroid tissues from THCA patients was examined by RT-qPCR. Data were normalized to GAPDH. ∗∗∗p < 0.001. E, representative images showing UBE2I protein expression in 43 pairs of tumor tissues and matched adjacent non-tumorous thyroid tissues from THCA patients by Western blot. F, the relative protein expression of UBE2I in THCA cell lines was determined by Western blot. Data were normalized to β-actin. ∗∗∗p < 0.001 vs. Nthy-ori 3-1.

UBE2I promoted THCA cell proliferation, migration, and invasion

High expression of UBE2I in THCA suggested a potential function for UBE2I in the development of THCA. To investigate the role of UBE2I in the malignant processes of THCA, UBE2I was upregulated or downregulated in CAL62 and TPC-1 cells, respectively (Fig. 2A). Upregulation of UBE2I enhanced proliferation, whereas knockdown of UBE2I notably diminished proliferation in both cell lines (Fig. 2, B and C). Besides, overexpression of UBE2I significantly increased the numbers of migratory and invasive cells, while silencing of UBE2I markedly decreased these numbers in both cell lines (Fig. 2D). Meanwhile, E-cadherin expression was significantly decreased, whereas Vimentin expression was increased in both cell lines overexpressing UBE2I (Fig. 2E). On the other hand, opposite changes in E-cadherin and Vimentin expression were observed in both cell lines with UBE2I downregulation (Fig. 2E). Altogether, these results indicate that UBE2I promotes THCA cell proliferation, migration, and invasion.

Figure 2.

UBE2I accelerated THCA cell proliferation, migration and invasion. shUBE2I was transfected into CAL62 and TPC-1 cells to downregulate the expression level of UBE2I, while UBE2I sequences were inserted into pcDNA vector plasmids and then transfected into CAL62 and TPC-1 cells to overexpress the expression level of UBE2I. A, the knockdown or overexpression efficiency was determined by Western blot in CAL62 and TPC-1 cells. Data were normalized to β-actin. B, cell viability of CAL62 and TPC-1 cells was determined by CCK-8 assays. C, the proliferation of CAL62 and TPC-1 cells was examined by EdU assays. The Y-axis represents the percentage of EdU-positive cells per field. Scale bar = 100 μm. D, the migratory and invasive abilities of CAL62 and TPC-1 cells were detected by transwell assays. E, the relative protein expression of E-cadherin and Vimentin in CAL62 and TPC-1 cells was determined by Western blot. Data were normalized to β-actin. ∗p < 0.05, ∗∗p < 0.01 and ∗∗∗p < 0.001 vs. NC; #p < 0.05, ##p < 0.01 and ###p < 0.001 vs. shNC. The y-axis is truncated to better visualize the differences between experimental groups.

UBE2I induced migration and invasion and M2 polarization of macrophages

To investigate the immunological relevance of UBE2I in thyroid cancer, we next assessed its impact on macrophage migration and polarization using a Transwell assay and conditioned medium (CM) from THCA cells. The data showed that UBE2I levels were positively correlated with macrophages (Fig. 3A). To further clarify the role of UBE2I in macrophage migration, macrophages were seeded in the upper chamber of a Transwell plate and incubated with CM from TPC-1 and CAL62 cells for 10 h. Macrophage migration induced by CM from cells transfected with NC or empty vector was clearly increased compared with that induced by control CM (Fig. 3B). This effect was further increased with CM from both cells transfected with UBE2I overexpression plasmid, whereas it was reduced with CM from cells transfected with shUBE2I (Fig. 3B). Moreover, the levels of M2 macrophage markers (CD206 and CD163) and the M2 cytokine IL10 were prominently increased in macrophages treated with CM from cells transfected with NC or empty vector compared with control CM (Fig. 3, C and D). Similarly, these increases were significantly elevated with CM from TPC-1 and CAL62 cells after UBE2I overexpression, while they were observably reduced with CM from TPC-1 and CAL62 cells after shUBE2I transfection (Fig. 3, C and D). Taken together, these results indicate that UBE2I enhances macrophage migration and promotes M2 polarization.

Figure 3.

UBE2I enhanced migration and M2 polarization of macrophages.A, correlation between UBE2I expression and immune cell infiltration levels in thyroid cancer based on TIMER database analysis. The left-most panel (“Purity”) indicates the association between UBE2I expression and tumor purity, where a positive correlation suggests that UBE2I is mainly expressed in tumor cells rather than in the microenvironment. The remaining panels show the correlations between UBE2I expression and infiltration of six immune cell types: B cells, CD4⁺ T cells, CD8⁺ T cells, neutrophils, macrophages, and dendritic cells. B, migration of macrophages cultured with different CM. C, the relative protein expression of CD206 and CD163 was determined by Western blot. Data were normalized to β-actin. D, the concentrations of IL-10 were measured by ELISA. ∗∗∗p < 0.001 vs. control; #p < 0.05, ##p < 0.01 and ###p < 0.001 vs. NC + shNC CM.

UBE2I promoted the localization of hnRNPA2B1 in the cytoplasm by enhancing SUMOylation

To explore the mechanism of UBE2I in THCA progression, we examined the role of UBE2I in SUMOylation. Several studies have reported that hnRNPA2B1 is a target of UBE2I-mediated SUMOylation (29, 30). We examined the effect of UBE2I on the SUMOylation of hnRNPA2B1. Co-immunoprecipitation assays revealed that UBE2I overexpression markedly enhanced the conjugation of SUMO1 to hnRNPA2B1, indicating that hnRNPA2B1 is a SUMOylation substrate of UBE2I (Fig. 4A). Treatment with the SUMOylation inhibitor anacardic acid (AA) effectively reduced the interaction between SUMO1 and hnRNPA2B1, as verified by reciprocal IP experiments (Fig. 4B). Immunofluorescence staining further showed that UBE2I overexpression promoted the cytoplasmic accumulation of hnRNPA2B1, whereas AA treatment reversed this change in localization (Fig. C). As shown in Fig. S1, UBE2I overexpression markedly increased the level of SUMO1-conjugated hnRNPA2B1, whereas AA treatment efficiently reduced this modification. The input blots also revealed a general reduction in SUMO1-conjugated proteins after AA exposure, confirming its inhibitory effect on global SUMOylation. Moreover, Co-IP analysis using SUMOylation-deficient hnRNPA2B1 mutants confirmed that mutation of SUMOylation sites abolished the interaction with SUMO1 (Fig. 4D). Together, these results demonstrate that UBE2I enhances the SUMOylation of hnRNPA2B1 and promotes its cytoplasmic localization in thyroid cancer cells.

Figure 4.

UBE2I facilitated the localization of hnRNPA2B1 in the cytoplasm by promoting SUMOylation.A, co-immunoprecipitation (Co-IP) analysis showing that UBE2I overexpression enhanced SUMO1 conjugation of hnRNPA2B1 in TPC-1 cells. SUMOylated hnRNPA2B1 (s-hnRNPA2B1) was detected at ∼36 kDa, while the free SUMO1 band appeared at ∼12 kDa. B, Co-IP assays confirmed that treatment with the SUMOylation inhibitor anacardic acid (AA, 100 μM) markedly reduced hnRNPA2B1 SUMOylation induced by UBE2I overexpression. Reciprocal IP using an anti-SUMO1 antibody further supported the reduced interaction between SUMO1 and hnRNPA2B1 upon AA treatment. C, immunofluorescence staining of hnRNPA2B1 (green) showing that UBE2I overexpression promoted cytoplasmic localization of hnRNPA2B1, which was reversed by AA treatment. Nuclei were counterstained with DAPI (blue). Scale bar = 50 μm. D, Co-IP assay demonstrating that the SUMOylation-deficient hnRNPA2B1 mutant (MUT) failed to interact with SUMO1 compared with wild-type (NC), confirming that UBE2I-mediated SUMOylation occurs at specific lysine residues of hnRNPA2B1.

The carcinogenesis of UBE2I was reversed by the inhibition of SUMOylation modification

To further investigate the role of UBE2I-mediated SUMOylation in THCA progression, the SUMOylation inhibitor, AA was used to treat TPC-1 cells. Overexpression of UBE2I markedly increased cell viability, the numbers of migratory and invasive cells, macrophage migration, and IL-10 concentrations, and these effects were substantially neutralized by AA treatment (Fig. 5, A and B, and D). As shown in Figure 5C, conditioned medium (CM) from UBE2I-overexpressing cells (UBE2I CM) significantly enhanced macrophage migration compared with control or NC+shNC CM, suggesting that UBE2I upregulation in tumor cells promotes the secretion of soluble factors that attract macrophages. In contrast, AA treatment notably reduced macrophage migration in both the AA CM and UBE2I+AA CM groups, indicating that inhibition of SUMOylation can counteract the pro-migratory effect induced by UBE2I. Hence, these findings indicate that inhibition of SUMOylation modification abrogated the carcinogenic effects of UBE2I in THCA.

Figure 5.

The carcinogenesis of UBE2I was reversed by the inhibition of SUMOylation modification.A, cell viability of CAL62 and TPC-1 cells was determined by CCK-8 assays. B, the migratory and invasive ability of CAL62 and TPC-1 cells was detected by transwell assays. C, Transwell migration assays showing the effect of conditioned medium (CM) from thyroid cancer cells under different treatments—control, NC + shNC CM, UBE2I CM, AA CM, and UBE2I + AA CM—on THP-1-derived macrophages. CM from UBE2I-overexpressing cells markedly enhanced macrophage migration, whereas treatment with the SUMOylation inhibitor anacardic acid (AA, 100 μM) significantly reduced this effect. Representative images and quantification of migrated cells are shown. D, the concentrations of IL-10 were measured by ELISA. ∗∗p < 0.01 vs. Control; ##p < 0.01 vs. NC + shNC CM; &&p < 0.01 vs. UBE2I CM.

Silencing of UBE2I suppressed the cell growth of THCA in tumorous mice

Moreover, the role of UBE2I was examined in vivo after mice were xenografted with TPC-1 cells transfected with shUBE2I. Knockdown of UBE2I markedly reduced tumor weight and size (Fig. 6A), as well as the levels of Ki-67 and CD206 (Fig. 6B). In addition, UBE2I expression was significantly decreased in UBE2I-depleted tumors at the 28-days time point (Fig. 6C). To assess the impact of UBE2I on tumor-associated macrophages (TAMs) in vivo, we performed immunofluorescence co-staining of the macrophage marker F4/80 and the M2 marker CD206 in xenograft tumors. Knockdown of UBE2I (shUBE2I) significantly reduced both total macrophage infiltration (F4/80+) and M2-polarized macrophages (CD206+) compared with the control group (shNC) (Fig. 6D). Overall, silencing of UBE2I restrained THCA growth in vivo.

Figure 6.

Downregulation of UBE2I inhibited the growth of THCA in vivo. Nude mice were inoculated in the right flank with 5 × 10ˆ6 TPC-1 cells transfected with shUBE2I or shNC. Tumor volume was monitored every 7 days for four consecutive weeks and calculated using the formula: volume = half × length × widthˆ2. After 24 days, mice were sacrificed, and tumor samples were excised, weighed, and fixed in 4% formaldehyde for immunohistochemistry assays. A, representative images of tumors from nude mice (upper), tumor volume curves (middle), and tumor weight (lower). B, the expression levels of Ki-67 and CD206 were detected by immunohistochemistry. C, the expression of UBE2I in tumor tissues from the indicated groups at the 28-day time point. D, immunostaining showing the expression of F4/80 (red) and CD206 (green) in tumor tissues from the indicated groups. ∗∗∗p < 0.001 vs. shNC.

UBE2I promotes THCA progression and macrophage M2 polarization through hnRNPA2B1 SUMOylation

To further elucidate the mechanistic role of hnRNPA2B1 SUMOylation in UBE2I-mediated THCA progression, we examined the SUMOylation status of hnRNPA2B1 wild-type (WT) and a SUMOylation-deficient mutant (MUT) in TPC-1 cells. Co-IP assays revealed that UBE2I overexpression significantly increased SUMO1 conjugation to hnRNPA2B1 WT, but not to the MUT (Fig. 7A). Functional assays demonstrated that the UBE2I-driven increases in migration and invasion of TPC-1 cells depended on hnRNPA2B1 SUMOylation, since these effects were abolished in cells expressing the MUT (Fig. 7B). As shown in Fig. S2, co-overexpression of UBE2I and hnRNPA2B1WT markedly increased cytoplasmic accumulation of hnRNPA2B1, while cells expressing the SUMOylation-deficient mutant showed mainly nuclear localization. In addition, CM from TPC-1 cells overexpressing UBE2I and hnRNPA2B1 WT markedly promoted macrophage migration and IL-10 secretion, while CM from cells expressing hnRNPA2B1 MUT did not produce these effects (Fig. 7, C and D). As shown in Figure 7E, co-expression of UBE2I and hnRNPA2B1 WT markedly enhanced SUMO1 conjugation of hnRNPA2B1, whereas this modification was nearly absent in cells expressing the SUMOylation-deficient mutant. Together, these results indicate that UBE2I promotes THCA malignancy and macrophage M2 polarization by enhancing hnRNPA2B1 SUMOylation, which facilitates its cytoplasmic localization and supports its tumor-promoting functions.

Figure 7.

UBE2I promotes THCA progression and macrophage M2 polarization through hnRNPA2B1 SUMOylation.A, SUMOylation levels of hnRNPA2B1 wild-type (WT) and mutant (MUT) in TPC-1 cells with or without UBE2I overexpression, detected by Co-IP. B, migratory and invasive abilities of TPC-1 cells transfected with hnRNPA2B1 WT or MUT under UBE2I overexpression. C, migration of macrophages cultured with conditioned medium (CM) from TPC-1 cells expressing hnRNPA2B1 WT or MUT with UBE2I overexpression. D, IL-10 secretion by macrophages treated with CM from TPC-1 cells expressing hnRNPA2B1 WT or MUT with UBE2I overexpression. E, Western blot analysis of SUMO1 conjugation in TPC-1 cells co-transfected with UBE2I and either hnRNPA2B1 WT or the SUMOylation-deficient mutant (hnRNPA2B1 MUT). Data are presented as mean ± SD. ∗∗p < 0.01, ∗∗∗p < 0.001 vs. NC; ##p < 0.01, ###p < 0.001 vs. UBE2I; &&p < 0.01 vs. hnRNPA2B1 WT.

Discussion

THCA is a malignant neoplasm that has gained increasing attention in recent years due to its rising global incidence (2). PTM by SUMOylation has been involved in the THCA (31). UBE2I, an enzyme required for SUMO conjugation, is crucial to the development of different malignancies (26, 28, 32). In this study, we found that UBE2I was upregulated in THCA. Overexpression and knockdown experiments showed that UBE2I promoted cell growth, migration, and invasion, as well as M2 macrophage polarization in THCA cells. Mechanistically, UBE2I enhanced the localization of hnRNPA2B1 in the cytoplasm by promoting its SUMOylation, and inhibition of SUMOylation reversed the tumor-promoting effects of UBE2I overexpression in THCA cells. In vivo, downregulation of UBE2I reduced tumor growth and decreased the level of CD206. Taken together, these findings suggest that UBE2I promoted growth, migration, invasion, and M2 macrophage polarization in THCA cells by regulating the nuclear translocation of hnRNPA2B1 via SUMOylation.

UBE2I is considered an oncogene because it is upregulated in various tumors, including pan-digestive system tumors (33), hepatocellular carcinoma (28), oral squamous cell carcinoma (34), clear cell renal cell carcinoma (35), and glioblastoma (36). Consistent with these reports, we also found increased UBE2I expression in both THCA tissues and cells in this study, suggesting that UBE2I may contribute to THCA development. Yang et al. (26) reported that UBE2I knockdown significantly inhibited the migration and invasion of HCC cells. Huang et al. (28) showed that UBE2I silencing suppressed tumorigenesis and increased chemosensitivity in cholangiocarcinoma. UBE2I also mediates the pro-tumor effects of miR-10a-5p in cervical cancer, including metastasis, apoptosis, and cell growth (32). Similarly, our results showed that UBE2I promoted THCA cell proliferation, as confirmed by both CCK-8 and EdU assays. In addition, UBE2I increased the number of migratory and invasive THCA cells, accompanied by reduced E-cadherin and increased Vimentin levels. Moreover, UBE2I knockdown reduced tumor growth in tumor-bearing mice. Collectively, these findings indicate that UBE2I is overexpressed in THCA and promotes cell growth, migration, and invasion.

As one of the most promising approaches for cancer treatment, immunotherapy has evolved into the modern age of cancer therapeutics (37). Immune-evading cancers share several key features, including the ability of tumor cells to drive tumor-infiltrating leukocytes from a pro-inflammatory to an anti-inflammatory state, thereby preventing effective tumor elimination. Tumor-associated macrophages (TAMs), obtained through circulating monocyte progenitors, participate in tumor progression, suppress anti-tumor immune responses, and promote neovascular formation (38, 39). In solid tumors, high TAM infiltration is typically linked to a poor prognosis (40, 41). Also, TAMs have been reported as potential therapeutic targets in THCA (42). Macrophages can polarize into an M1 type, which is generally activated and shows anti-tumor effects, or an M2 type, which exhibits anti-inflammatory features but may promote tumor development (43). Accordingly, a major focus of current tumor immunotherapy is to promote TAM polarization toward M1 while reducing M2 polarization (44, 45, 46). Qin et al. (47) identified IRX5 as an oncogene in THCA, showing that IRX5 promotes M2 polarization and reduces M1 polarization of macrophages. Similarly, Cdc42 contributes to tumor-related M2 macrophage polarization and may represent a potential therapeutic target in THCA (48). In line with these findings, our work revealed that UBE2I expression was positively correlated with macrophages and that UBE2I enhanced macrophage migration as well as the levels of CD206, CD163, and IL-10. CD206 and CD163 are macrophage markers, and IL-10 is an M2-associated cytokine that can be used to evaluate M2 polarization (49). Therefore, these results suggest that UBE2I promotes macrophage migration and M2 polarization, supporting the view that UBE2I may be a potential immunotherapy-related target in THCA.

Most SUMO is found in nuclei as it is necessary to perform a number of nuclear processes, including directing the course of the cell cycle, controlling nucleocytoplasmic trafficking, responding to DNA damage, and modulating gene expression (50, 51). It is widely known that SUMOylation affects cargo protein properties and controls the function of transport apparatus components during nucleocytoplasmic transport (52, 53, 54). It has been demonstrated that endogenous SUMO-1 modification is involved in the platelet-derived growth factor-C’s nuclear localization in thyroid tissue and cell lines (31). In the current study, UBE2I enhanced the localization of hnRNPA2B1 in the cytoplasm by promoting its SUMOylation. One member of the heterogeneous nuclear ribonucleoproteins, hnRNPA2B1, is mostly found in the nucleus (55). HnRNPA2B1 is involved in a wide range of biological activities in cancers (56). Silencing of hnRNPA2B1 has been shown to inhibit cervical cancer cell growth and invasion (57), and hnRNPA2B1 promotes epithelial-mesenchymal transition in head and neck tumor cells (58). Moreover, hnRNPA2B1 SUMOylation is closely related to the pathologies of amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD) (59). SUMOylated hnRNPA2B1 tends to accumulate in the cytoplasm, where it can form harmful protein aggregates that contribute to neuronal dysfunction and death (59). In addition, SUMOylation of hnRNPA2B1 is increased by HBV infection, which facilitates its translocation from the nucleus to the cytoplasm, thereby supporting HBV RNA transport and translation and promoting viral replication (60). A previous study has also reported that patients with diabetic nephropathy have higher levels of SUMOylated hnRNPA2B1, which leads to cytoplasmic accumulation of the protein (61). Since hnRNPA2B1 SUMOylation is linked to kidney cell injury and fibrosis, regulating this modification may provide a potential treatment strategy for diabetic nephropathy (61). Consistently, Xu et al. demonstrated that UBE2I regulates hnRNPA2B1 nuclear translocation in osteoarthritis by enhancing SUMO modification (30).

hnRNPA2B1 is also known to play roles in RNA metabolism, subcellular trafficking, and stress granule formation, all of which have been implicated in cancer progression. Based on this background, we hypothesized that SUMOylation of hnRNPA2B1 might mediate at least part of the downstream effects of UBE2I in thyroid cancer. While our current findings support the SUMOylation and cytoplasmic redistribution of hnRNPA2B1, we agree that further validation is needed to confirm the functional significance of this change. This study also showed that inhibition of SUMOylation abrogated the carcinogenic effects of UBE2I in THCA. Overall, these results suggest that UBE2I promotes tumor progression and immune infiltration in THCA, at least in part, through hnRNPA2B1 SUMOylation. Our initial bioinformatic analysis revealed a positive correlation between UBE2I expression and macrophage infiltration in thyroid cancer, which prompted us to explore the immunological impact of UBE2I. Although UBE2I appears to be primarily dysregulated in tumor cells, our results suggest that it may exert indirect immunomodulatory effects by altering the tumor cell secretome that drives macrophage migration and M2 polarization. Further studies are needed to clarify the key molecular mediators involved in this process. Correlation analysis using the TIMER database also revealed that UBE2I expression is positively associated with tumor purity, suggesting predominant expression in tumor cells. Interestingly, despite this tumor-intrinsic pattern, UBE2I was also positively correlated with macrophage infiltration, whereas associations with other immune cell types were less consistent. This selective relationship, together with the established role of tumor-associated macrophages in promoting thyroid cancer progression, prompted us to investigate macrophage behavior in response to tumor cell-derived signals modulated by UBE2I.

Notably, a recent study also reported that UBC9 (UBE2I) is upregulated in papillary thyroid carcinoma and promotes tumor proliferation and migration. While these findings support a pro-tumor role of UBE2I, our study extends this work by providing additional mechanistic insights into how UBE2I exerts its effects (30). Specifically, we show that UBE2I promotes the SUMOylation and cytoplasmic localization of hnRNPA2B1, a pathway that was not explored in the prior study. Additionally, our study is the first to demonstrate that UBE2I can influence the tumor immune microenvironment by promoting M2 macrophage polarization. This dual role, affecting both cancer cell behavior and immune infiltration, highlights UBE2I as a multifaceted regulator of thyroid cancer progression and suggests a potential rationale for therapeutic strategies targeting both tumor cells and the immune microenvironment.

Although our current study demonstrates that UBE2I promotes SUMOylation and cytoplasmic translocation of hnRNPA2B1 in thyroid cancer cells, we acknowledge that the direct functional role of hnRNPA2B1 in THCA has not yet been fully established. Previous reports have implicated hnRNPA2B1 in several malignancies, including cervical cancer, head and neck tumors, and multiple myeloma, where it contributes to proliferation, invasion, and epithelial–mesenchymal transition. Moreover, SUMOylated hnRNPA2B1 has been linked to pathological processes such as viral replication and diabetic nephropathy, which further supports its capacity to alter subcellular localization and influence disease progression. Based on this evidence, we hypothesize that hnRNPA2B1 may serve as an important mediator of UBE2I-driven tumor progression in THCA. Future studies using SUMOylation-deficient hnRNPA2B1 mutants together with in vivo functional assays will be essential to determine whether hnRNPA2B1 is not only a substrate of UBE2I but also a functional downstream driver of thyroid cancer progression.

Although we demonstrated that UBE2I promotes the SUMOylation and cytoplasmic translocation of hnRNPA2B1, the direct functional impact of this modification on thyroid cancer cell proliferation, migration, or immune modulation remains to be fully clarified. Future studies involving SUMOylation-deficient hnRNPA2B1 mutants will be essential to establish a causal link between hnRNPA2B1 SUMOylation and UBE2I-mediated tumor progression.

While our in vitro results indicate that UBE2I may enhance macrophage migration and M2 polarization, we acknowledge that these findings alone do not fully establish the physiological relevance of UBE2I-mediated immune modulation in vivo. Although CD206 expression was reduced in tumor tissues following UBE2I knockdown, future studies using flow cytometry and immunofluorescence co-staining with macrophage-specific markers will be needed to define macrophage phenotype and function more clearly within the tumor microenvironment. In parallel, studies using SUMOylation-deficient hnRNPA2B1 mutants and in vivo models could be critical in determining whether hnRNPA2B1 acts as a bona fide downstream effector of UBE2I in thyroid cancer. Collectively, these investigations will help clarify whether the immunomodulatory effects observed in vitro translate into functional immune alterations in vivo.

In addition, although our in vitro findings suggest that UBE2I may contribute to macrophage polarization toward an M2 phenotype, we recognize that this effect is likely part of a broader signaling network and cannot be attributed solely to UBE2I overexpression. Accordingly, further in vivo work with co-staining and flow cytometry will be important to validate the functional phenotype of macrophages in the THCA microenvironment.

In conclusion, UBE2I expression was elevated in THCA, and UBE2I promoted proliferation, migration, invasion, and M2 macrophage polarization through hnRNPA2B1 SUMOylation in THCA. Nevertheless, the relationship between UBE2I levels and clinically important pathological parameters in THCA patients should be explored to support potential clinical applications. Likewise, the role of UBE2I in other malignant processes in THCA, such as epithelial–mesenchymal transition and apoptosis, warrants further investigation. Finally, a more comprehensive evaluation of UBE2I function and its mechanisms in THCA progression should be carried out in vivo. Overall, these findings identify a potential target for THCA detection and therapy.

Experimental procedures

The UBE2I expression profile in THCA in silico

GEPIA (http://gepia2.cancer-pku.cn) and UALCAN (http://ualcan.path.uab.edu/index.html) were utilized to examine the UBE2I expression profile in para-carcinoma normal samples and THCA samples, respectively. In addition, the GSE129562 microarray dataset, which was obtained from the GEO database and based on the GPL10558 Illumina HumanHT-12 V4.0 platform, included eight thyroid tissues from patients with thyroid papillary microcarcinoma who underwent thyroidectomy and eight normal thyroid tissues from the same patients’ thyroid lobes. LIMMA software in R was employed to determine UBE2I expression in the GSE129562 microarray dataset (62).

Tissue sample

Patients with THCA who were diagnosed and treated in our hospital provided 43 pairs of tumor tissues and matched adjacent non-tumorous thyroid tissues, which were immediately placed in liquid nitrogen until further procedures. Before surgery, no patient received preoperative radiotherapy, chemotherapy, or immunotherapy. All patients had only one type of cancer. The normal tissues adjacent to the tumor were separated by at least 3 cm. This work was approved by the Board and Ethics Committee of Tianjin Union Medical Center and was performed following the Helsinki Declaration. All participants were informed about the study details and provided signed consent (Approval No.2022-C18).

RT-qPCR

Following a previous report (63), TRIzol reagent (15596026, Thermo Fisher Scientific) was employed to extract total RNA from THCA tissues. The Bio-Rad ScripTM cDNA Synthesis Kit (1,708,890, Bio-Rad, Inc) was then used to perform reverse transcription according to the manufacturer’s instructions. The 2× SYBR Master mix (RR820A, Takara, Dalian, China) was used to perform RT-qPCR on the Bio-Rad CFX Manager system. UBE2I expression was calculated using the 2^-ΔΔCT method with GAPDH as the internal control. The primer sequences were as follows: 5′-AAAAATCCCGATGGCACGATG-3′ (UBE2I forward), 5′-CTTCCCACGGAGTCCCTTTC-3′ (UBE2I reverse); 5′-GGTGGTCTCCTCTGACTTCAACA-3′ (GAPDH forward), 5′-GTTGCTGTAGCCAAATTCGTTGT-3′ (GAPDH reverse).

Cell culture

The THCA cell lines IHH-4 (CL-0803), CAL62 (CL-0618), and TPC-1 (CL-0643), as well as the human thyroid follicular epithelial cell line Nthy-ori 3-1 (CL-0817), were obtained from Procell. Cells were cultured in RPMI-1640 medium (PM150110, Procell) supplemented with 10% fetal bovine serum (FBS, 164,210, Procell) and 1% penicillin-streptomycin (PB180120, Procell), and maintained at 37°C in a 5% CO₂ incubator.

Cell transfection and treatment

GenePharma designed and provided short-hairpin RNAs (shRNAs) targeting UBE2I (shUBE2I) and a scrambled shRNA control (shNC). In addition, the UBE2I sequence was inserted into pcDNA vector plasmids to upregulate UBE2I expression. Transfection of CAL62 and TPC-1 cells was performed using Lipofectamine 3000 (L3000001, Invitrogen). The knockdown or overexpression efficiency was assessed by Western blotting. To evaluate the role of SUMOylation, TPC-1 cells were treated with 50 μM of the SUMOylation inhibitor anacardic acid (AA) for 72 h based on a previous study (20). AA (98.86%) was purchased from MedChemExpress (HY-N2020, Monmouth Junction). AA was dissolved in DMSO (D8371, Solarbio) and then diluted in PBS (P1020, Solarbio). This approach was chosen to simulate the paracrine communication between THCA cells and macrophages, thereby enabling assessment of tumor cell–derived soluble factors on macrophage behavior.

Cell counting kit-8

TPC-1 and CAL62 cells were seeded into 96-well plates at a density of 5 × 10³ cells/well and cultured at 37 °C in a 5% CO₂ incubator. Then, 10 μl of CCK-8 reagent (CA1210, Solarbio) was added to each well and incubated at 37 °C for 2 h. Absorbance at 450 nm was measured using a microplate reader (Thermo Fisher Scientific).

five-Ethynyl-2′-deoxyuridine (EdU) incorporation assay

EdU assays were performed using EdU kits with Alexa Fluor 594 (C0078S, Beyotime). Briefly, TPC-1 and CAL62 cells were seeded and cultured for 12 h. Each well was then supplemented with 1 ml of 20 μM EdU solution, followed by incubation for 2 h at 37 °C. Cells were subsequently incubated with anti-EdU Click reaction solution for 30 min in the dark. Afterwards, cells were fixed with fixing solution (P0098, Beyotime) and permeabilized with 0.3% Triton X-100 (T8200, Solarbio). For nuclear staining, 5 μg/ml Hoechst 33,342 (P0133, Beyotime) was used. The cells were observed using a fluorescent microscope (Olympus). The proportion of EdU-positive cells was determined by randomly selecting five fields.

Transwell assay

For the invasion assay, 200 μl of TPC-1 or CAL62 cell suspension (5 × 10⁴ cells) was seeded into the upper chamber of a 24-well transwell plate (8-μm pores, #3422, Corning) pre-coated with Matrigel (#356234, Solarbio). For the migration assay, cell migration was evaluated using a similar transwell setup without Matrigel coating, and 200 μl of cell suspension (5 × 10⁴ cells) was added to the uncoated upper chamber. The lower chamber was filled with RPMI-1640 medium supplemented with 20% fetal bovine serum (FBS). After 48 h of incubation, the Matrigel and any non-invading cells on the upper surface of the membrane were carefully removed with a cotton swab. Cells that had invaded through the Matrigel and membrane were fixed with 4% paraformaldehyde (P1110, Solarbio), stained with 0.1% crystal violet (G1063, Solarbio), and imaged using an inverted microscope. The number of invaded cells was quantified by randomly counting cells in five distinct fields per well.

Immune infiltrate analyses in THCA

The Tumor Immune Estimation Resource (TIMER; https://cistrome.shinyapps.io/timer/) database was used to examine immune infiltrates in THCA. In particular, the gene module was used to analyze the correlation between UBE2I expression and the abundance of immune infiltrates, including neutrophils, dendritic cells, macrophages, CD4+ T cells, CD8+ T cells, and B cells. The relationships between UBE2I expression and immune infiltrates were assessed using Spearman’s correlation coefficients.

The polarization model of macrophages

THP-1 cells were stimulated with phorbol 12-myristate 13-acetate (PMA, 99.80% purity, HY-18739, MedChemExpress) to differentiate into M0 macrophages. The M0 macrophages were then activated with interleukin 4 (IL-4, SRP3093, Sigma-Aldrich) and IL-13 (SRP3274, Sigma-Aldrich) to induce polarization into M2 macrophages.

Preparation of conditioned medium (CM)

After seeding 5 × 10⁴ CAL62 and TPC-1 cells per well in 12-well plates for 24 h, the supernatants were collected. Four types of CM were prepared for macrophage incubation, including control CM, NC + shNC CM, UBE2I CM, and shUBE2I CM. Specifically, shUBE2I CM was collected from CAL62 and TPC-1 cells transfected with shUBE2I, UBE2I CM was collected from CAL62 and TPC-1 cells overexpressing UBE2I, NC + shNC CM was collected from CAL62 and TPC-1 cells transfected with shNC and empty vector, and control CM was collected from culture medium without tumor cells.

Macrophage migration assay

After in vitro polarization of THP-1 cells, 6.5 mm transwell plates with 5.0 μm pores (09,717,050, Corning Company) were used for the migration assay. Macrophages were seeded into the upper chamber at a density of 1 × 10ˆ5 cells/well, and the lower chamber was filled with CM. After 10 h, the cell suspension in the upper chamber was removed, and the upper surface of the membrane was gently cleaned with cotton swabs. Macrophages were stained with crystal violet, and five representative fields per membrane were imaged. Migratory cells on the lower surface were counted using ImageJ and analyzed statistically.

Enzyme-linked immunosorbent assay (ELISA)

After macrophages were cultured with different types of CM, the culture supernatants were centrifuged at 1000×g for 15 min to remove cell debris before ELISA. The Human IL-10 ELISA Kit (PI536, Beyotime) was used to determine IL-10 levels according to the manufacturer’s instructions.

Co-immunoprecipitation (Co-IP) assay

To obtain cell lysates, TPC-1 cells were lysed with RIPA lysis buffer (R0100, Solarbio) containing protease inhibitors (A8260, Solarbio) for 40 min on ice. After centrifugation, the supernatants were collected. Target antibodies or immunoglobulin G (IgG) as a negative control were added to the supernatants and incubated overnight at 4°C. Protein A/G-agarose beads (78,609, Thermo Fisher Scientific) were then added and gently rotated at 4°C. After incubation, the beads were collected and washed with lysis buffer. The immunoprecipitates were boiled in loading buffer (P1040, Solarbio) for 5 min and analyzed by Western blotting.

Western blotting

Following an earlier report (64), total proteins were extracted from tissue and cell samples using RIPA lysis buffer. Protein concentrations were measured using a BCA Protein Assay Kit (PC0020, Solarbio). Equal amounts of protein (20 μg) were separated by 10% SDS-PAGE and transferred onto PVDF membranes (IPVH00010, EMD Millipore, Billerica, MA, USA). After blocking with 5% BSA blocking buffer (SW3015, Solarbio), the membranes were incubated with primary antibodies at 4 °C overnight. The membranes were then incubated with secondary antibodies (1:5000, ab288151, Abcam) at room temperature for 1 h. Protein bands were visualized using the BeyoECL Plus kit (P0018S, Beyotime), and gray values were quantified using Image-ProPlus (Media Cybernetics, Inc., Rockville, MD, USA). β-actin served as an internal reference. The primary antibodies used were as follows: UBE2I (1:1000, ab33044, Abcam), β-actin (1:5000, ab8227), SUMO1 (1:2500, ab32058), hnRNPA2B1 (1:1000, ab31645), CD163 (1:1000, ab87099), CD206 (1:1000, ab64693), Vimentin (1:2500, ab92547), and E-cadherin (1:25,000, ab40772).

Immunofluorescence (IF) assay

TPC-1 cells were seeded on glass coverslips and cultured at 37 °C in a 5% CO₂ incubator. After washing three times with PBS, cells were fixed with 4% paraformaldehyde at room temperature for 15 min and washed again with PBS. Cells were permeabilized with 0.2% Triton X-100 and blocked with BSA blocking buffer. The cells were then incubated with primary antibodies against UBE2I (1:500, ab33044, Abcam), F4/80 (1:200, ab111101, Abcam), and CD206 (1:100, ab64693, Abcam), as indicated. After three PBS washes, cells were incubated with Goat Anti-Rabbit IgG H&L (Alexa Fluor 488) (1:500, ab150077, Abcam) at room temperature for 1 h. Finally, an antifading mounting medium with DAPI (S2110, Solarbio) was used for nuclear staining and mounting. Images were captured using an Olympus fluorescence microscope.

Animal experimentation

BALB/c nude mice (4 weeks old) were obtained from JKbiot (Nanjing, China) and maintained under specific pathogen-free (SPF) conditions with controlled temperature and a 12 h light/dark cycle. Mice were randomly assigned to the shNC group and the shUBE2I group (n = 6). TPC-1 cells were transfected with shUBE2I or shNC, and 5 × 10⁶ cells were subcutaneously injected into the right flank of nude mice in the corresponding groups (30). Tumor volume was measured every 5 days for four consecutive weeks and calculated using the formula: volume = 0.5 × length × widthˆ2. Mice were sacrificed after 4 weeks under isoflurane anesthesia (R510–22, RWD, Guangdong, China). Tumor tissues were excised, weighed, and stored for further assays. All animal procedures were approved by the Animal Research Ethics Committee of Nankai University and complied with the Guide for the Care and Use of Laboratory Animals (65).

Immunohistochemistry

Gradient ethanol was used to dehydrate tumor tissues following fixation with 4% paraformaldehyde. Afterwards, paraffin (YA0011, Solarbio) was utilized to embed the tissues, followed by slicing into 5 μm-thick sections. Antigen retrieval was performed with sodium citrate buffer (pH 6.0, P0081, Beyotime) and maintained for 15 min at 94 °C. After blocking with 1% BSA for 1 hour, the sections were treated with primary antibodies against Ki-67 (1:500, ab15580, Abcam) and CD206 (1:100, ab64693) at 4 °C overnight. Anti-rabbit IgG (ab288151, Abcam) as the secondary antibody was incubated with the sections at 37 °C for 30 min. The sections were examined under a light microscope (Olympus) after counterstaining with hematoxylin (G1080, Solarbio).

Statistical analysis

Data were analyzed using SPSS 20.0 (IBM) and are expressed as mean ± SD. One-way ANOVA or a t test, as well as the post hoc Bonferroni test, were employed to identify statistical variations among two or more groups. p < 0.05 was considered a significant difference.

Ethics approval

All procedures performed in studies involving human participants were in accordance with the standards upheld by the Ethics Committee of Tianjin Union Medical Center and with those of the 1964 Helsinki Declaration and its later amendments for ethical research involving human subjects (Approval No.2022-C18).

All animal experiments were approved by the Ethics Committee of Nankai University for the use of animals and conducted in accordance with the National Institutes of Health Laboratory Animal Care and Use Guidelines (Approval No.2022-SYDWLL-000218).

Data availability

Data created or determined in this work have been provided in published manuscript. Datasets employed and/or determined in current work would be provided by correspondence author upon acceptable demand.

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they do not have any conflicts of interest with the content of this article.

Reviewed by members of the JBC Editorial Board. Edited by Paul Shapiro