Abstract
Background
Uveal melanoma (UM) was prone to metastasis and had an extremely poor prognosis. High-mobility group protein A1 (HMGA1) was known to promote proliferation and invasion in various tumors, but its molecular mechanism in UM remained unclear.
Objective
The effect of HMGA1 on the proliferation, migration, and invasion of UM cells was investigated, and whether it functioned through the phosphatidylinositol-3-kinase (PI3K)/protein kinase B (Akt)/matrix metalloproteinase-9 (MMP-9) pathway was explored.
Methods
HMGA1 overexpression (OE) and knockdown (KD) models were established in two UM cell lines (C918, MUM-2B). Cell functions were assessed using MTT assay, EdU (5‑ethynyl‑2’‑deoxyuridine) incorporation assay, scratch wound healing assay, and Matrigel-Transwell assay. The expression of PI3K, phosphorylated PI3K (p-PI3K), Akt, phosphorylated Akt (p-Akt), and MMP-9 was detected by Western blot (WB). Pathway intervention was performed using the PI3K inhibitor LY294002.
Results
Compared with the control group, HMGA1 overexpression significantly up-regulated the proliferation, migration, and invasion abilities of both cell lines, and led to a fold-dependent up-regulation of PI3K/Akt pathway activation and MMP-9 expression. In contrast, HMGA1 knockdown significantly down-regulated the above indicators. After treatment with LY294002, the tumor-promoting effects induced by HMGA1 overexpression were significantly reversed.
Conclusion
HMGA1 significantly promoted the proliferation, migration, and invasion of UM cells from different origins (primary C918, metastatic MUM-2B) by activating the PI3K/Akt pathway and upregulating MMP-9 expression, suggesting its potential as a target for molecular targeted therapy in UM.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12920-026-02387-6.
Keywords: Uveal melanoma, HMGA1, PI3K/Akt pathway, MMP-9
Background
Uveal melanoma (UM) is the most common primary intraocular malignancy in adults, with an annual incidence of approximately 0.5-1/10,000 [1, 2]. Although significant progress has been made in local control (e.g., enucleation, radiotherapy), approximately 50% of patients develop distant metastasis after diagnosis, with liver metastasis being the most common [3, 4]. The survival of patients with metastasis is severely limited, and the prognosis is extremely poor. Currently, systemic treatment options for metastatic UM are limited [5]. Molecular targeted therapy and immunotherapy only provide short-term relief for a minority of patients. Therefore, it is urgent to explore the molecular drivers of UM to discover new therapeutic targets. HMGA1 is a non-histone chromatin remodeling factor that regulates chromatin spatial conformation and gene transcriptional activity by specifically binding to AT-rich DNA regions [6, 7]. HMGA1 is highly expressed during embryonic development but is expressed at very low levels in healthy adult tissues. Previous studies have confirmed that HMGA1 is highly expressed in various solid tumors and hematological malignancies, promoting cell proliferation, epithelial-mesenchymal transition, and extracellular matrix degradation [8, 9]. It has been shown that HMGA1 promotes malignant tumor progression by regulating the PI3K/Akt pathway [10]. HMGA1 can influence tumor cell proliferation and metastasis by modulating downstream transcription factors or signaling pathways [11]. In multiple tumor types, HMGA1 activates the PI3K/Akt/MMP-9 signaling axis [12, 13].
The PI3K/Akt pathway is a core pathway regulating cell proliferation, survival, and motility [14], and its aberrant activation is closely associated with the invasion and metastasis of uveal melanoma [15]. Upon activation, PI3K generates PIP3, promoting Akt phosphorylation and activating downstream effectors, thereby driving cell cycle progression and inhibiting apoptosis [16]. Akt signaling up-regulates various matrix metalloproteinases, among which MMP-9 plays a key role in tumor invasion and metastasis and is an important downstream effector molecule of the PI3K/Akt pathway. MMP-9-mediated matrix degradation creates conditions for tumor cells to penetrate the basement membrane and metastasize, making it a hotspot marker in many tumor studies. In particular, the development of UM is closely related to specific molecular alterations. Among them, GNAQ.Q2029L and GNA11.Q209L mutations are the core driver events in UM and can activate the PI3K/Akt pathway through downstream signaling cascades [17]. Additionally, PRAME overexpression, SF3B1 mutation, and BAP1 inactivation are also involved in the malignant progression of UM [18]. No study has yet elucidated whether HMGA1 regulates the malignant biological behavior of uveal melanoma cells through the PI3K/Akt/MMP-9 pathway, which is a scientific question urgently to be addressed in this study. Based on the above analysis, this study first proposed the hypothesis that HMGA1 upregulates MMP-9 expression by activating the PI3K/Akt signaling pathway, thereby promoting the proliferation, migration, and invasion of uveal melanoma cells. To test this hypothesis, this study employed gene overexpression and knockdown strategies, combined with functional assays and pathway intervention, to clarify the role of HMGA1 in the development of UM, providing a theoretical basis for molecular targeted therapy of UM.
Materials and methods
Cells and reagents
Uveal melanoma cell lines C918 and MUM-2B were purchased from the American Type Culture Collection (ATCC). STR genotyping confirmed consistency with the ATCC database, and mycoplasma contamination was ruled out using a mycoplasma detection kit (GIVEI).
Cell culture conditions: C918 cells were cultured in DMEM medium (SH30243.01, Cytiva), and MUM-2B cells were cultured in RPMI-1640 medium (SH30605.01, Cytiva). Both media were supplemented with 10% fetal bovine serum (FBS) (BC-SE-FBS01, SunBio) and 100 U/mL penicillin − 100 µg/mL streptomycin (RC20016, AMEKO). All cells were cultured in a 37 °C, 5% CO2 incubator, passaged at a 1:3 ratio, and the medium was changed every 3 days.
Antibodies: Anti-HMGA1 (HPA065612, Atlas Antibodies), anti-PI3K p85 (IPD00514, Aiptide, 1:1000), anti-p-PI3K (Tyr458) (CST 4228 L, Cell Signaling Technology, 1:1000), anti-Akt (FNab00271, FineTest, 1:1000), anti-p-Akt (Ser473) (WLP001a, Wanleibio, 1:1000), anti-MMP-9 (GB12132-100, Servicebio, 1:1000), anti-β-actin (bsm-33036 M-200µL, BIOSS, 1:5000).
Inhibitor: The PI3K inhibitor LY294002 (S1737-1 mg, Beyotime Biotechnology) was prepared in dimethyl sulfoxide (DMSO) as a 10 mM stock solution. The working concentration was determined through gradient experiments to be 20 µM, suitable for both cell lines.
siRNA and plasmids: HMGA1-siRNA sequences (si-1: 5’-GCAGUAUUUCGAGGCUAAATT-3’; si-2: 5’-UUAGCCUCGAAAAUACUGCTT-3’) and negative control siRNA (si-NC) were synthesized by Beijing Jinrui Baikang Biotechnology Co., Ltd. In this study, two separate siRNAs were used for transfection to verify the specificity and repeatability of the knockdown effect. Preliminary experiments confirmed that the knockdown efficiency of both siRNAs at both the mRNA and protein levels in the two cell lines was ≥ 70%. The overexpression vector was pcDNA3.1-HMGA1 (Shanghai Lianmai Bioengineering Co., Ltd., inserted sequence verified by sequencing), with empty pcDNA3.1 vector as the control.
Transfection: Lipofectamine 3000 (L3000001, Thermo Fisher Scientific) was used according to the manufacturer’s instructions. Transfection conditions were optimized: MUM-2B cells were transfected with 2.5 µg plasmid + 5 µL Lipofectamine 3000; C918 cells were transfected with 3 µg plasmid + 6 µL Lipofectamine 3000.
LY294002 dose screening experiment
To determine the optimal working concentration of LY294002, gradient concentrations of 0, 5, 10, 20, 40 µM were used to treat C918 and MUM-2B cells. After routine culture for 24 h, cells were collected (C918 cells proliferated at a slower rate and responded slightly slower to drug treatment, so the treatment duration was extended to 28 h), the expression levels of p-PI3K and p-Akt were detected by WB, and cell viability was assessed by MTT assay. The results showed that 20 µM LY294002 significantly reduced the p-PI3K/PI3K and p-Akt/Akt ratios in both cell lines, with an inhibition efficiency of over 50%, and cell viability remained above 70% without obvious cytotoxicity. At 40 µM, cytotoxicity increased significantly, with viability below 50%. Therefore, 20 µM was determined as the optimal working concentration.
HMGA1 gene operation
C918 and MUM-2B cells in the logarithmic growth phase were seeded into 6-well plates. Seeding density: MUM-2B: 2 × 10⁵ cells/well; C918: 2.5 × 10⁵ cells/well. Transfection was performed 24 h later when cell confluence reached 70%-80%. Overexpression: According to the Lipofectamine 3000 protocol, 2.5 µg (MUM-2B) or 3 µg (C918) pcDNA3.1-HMGA1 was mixed with the P3000 reagent, then combined with Lipofectamine 3000 dilution, incubated at room temperature for 15 min, and added to the cells. After 6 h, the medium was replaced with complete medium containing 10% FBS. G418 selection began 48 h post-transfection. The concentration was 800 µg/mL for C918 and 700 µg/mL for MUM-2B. Selection was maintained for two weeks until positive clones formed. Clones from each cell line were collected, and HMGA1 overexpression efficiency was verified by qRT-PCR and WB.
Under the same seeding conditions, 50 nM HMGA1-siRNA was transfected using Lipofectamine 3000. Total RNA was extracted using TRIzol 48 h post-transfection, and the reduction in HMGA1 mRNA level was detected by qRT-PCR (SYBR Green method), GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was used as the reference gene. Simultaneously, protein was extracted for WB detection to verify knockdown efficiency at the protein level. Only cells with knockdown efficiency ≥ 70% were used for subsequent experiments. Both overexpression and knockdown groups were set up with three biological replicates to compare changes in HMGA1 expression and phenotypic differences.
Functional experiments
Cell proliferation detection:
MTT assay: Transfected cells were seeded in 96-well plates, with 5 replicate wells per group. The seeding density was optimized: MUM-2B: 5 × 10³ cells/well; C918: 6 × 10³ cells/well. On days 1, 2, 3, and 4, 20 µL of 5 mg/mL MTT (298-93-1, Dixincheng Chemical) was added to each well and incubated at 37 °C for 4 h. The supernatant was removed, 150 µL DMSO was added, and the plate was shaken for 10 min. Absorbance (OD value) was measured at 570 nm. A proliferation curve was plotted with time as the X-axis and OD value as the Y-axis to assess the effect of HMGA1 overexpression/knockdown on proliferation in both cell lines.
EdU (5‑ethynyl‑2’‑deoxyuridine) incorporation assay: The Click-iT™ EdU kit (CSC6017S, Chemstan) was used. Cells were seeded at the density used in the MTT assay. After attachment, 10 µM EdU was added to each well, and both C918 and MUM-2B cells were incubated with 10 µM EdU for 3 h. Cells were washed 3 times with PBS (phosphate-buffered saline), fixed with 4% paraformaldehyde for 15 min, and permeabilized with 0.5% Triton X-100 for 10 min. Click reaction staining and DAPI nuclear counterstaining were performed according to the kit instructions. Five random fields per well were imaged under a fluorescence microscope. The percentage of EdU-positive cells to total cells (proliferation index) was calculated using ImageJ software.
-
(2)
Cell migration detection (scratch wound healing assay):
Cells were seeded in 6-well plates. The density was optimized: MUM-2B: 1 × 10⁶ cells/well; C918: 1.2 × 10⁶ cells/well, reaching 90% confluence within 24 h. A straight-line wound was created using a 200 µL pipette tip. After washing twice with PBS to remove detached cells, medium containing 2% FBS was added (3% FBS was used for C918 to maintain viability). Wound healing was monitored and photographed at 0, 24, and 48 h. Each group had 3 replicate wells, and 3 scratches were randomly selected per well. The width was measured using ImageJ software, and the wound closure rate was calculated: Closure rate (%) = (Initial width - Remaining width) / Initial width × 100%. The average closure rates of the two cell lines were compared.
-
(3)
Cell invasion detection (Transwell assay):
Matrigel-coated Transwell chambers (24-well, 8 μm pore size, Corning 3450) were used. The upper chamber was pre-coated with Matrigel diluted 1:8 (HYM 354234, Huayue Biotechnology) and solidified at 37 °C for 1 h. The lower chamber contained 600 µL of complete medium with 10% FBS as a chemoattractant. Transfected cells were suspended in serum-free medium and seeded into the upper chamber (200 µL/well): MUM-2B: 2 × 10⁵ cells/well; C918: 2.5 × 10⁵ cells/well. After incubation for 24 h (C918 cells for 30 h), non-invading cells were removed. Invaded cells were fixed with 4% paraformaldehyde for 15 min and stained with 0.1% crystal violet for 20 min. Five random fields per well were imaged under an inverted microscope. The average number of invaded cells was counted, and differences between groups were compared.
Signal pathway analysis
Protein extraction and WB detection:
Cells from each transfection group were collected, lysed with RIPA buffer containing phosphatase and protease inhibitors, incubated at 4 °C for 30 min, and centrifuged at 12,000 g for 15 min to collect the supernatant. Protein concentration was determined using the BCA method (PC0010, Solarbio) to ensure consistent sample concentration. 30 µg of protein per sample was separated by 10% SDS-PAGE (120 V, 90 min; extended to 100 min for C918 cells) and migrated to PVDF (polyvinylidene fluoride) membranes (66485, Lai Pu Biotechnology) at 100 V for 90 min. Membranes were blocked with 5% skim milk at room temperature for 1 h, incubated with primary antibodies (anti-PI3K p85, anti-p-PI3K, etc.) at 4 °C overnight. After washing three times with TBST buffer for 10 min each, HRP (horseradish peroxidase)-conjugated secondary antibody (QYL-030256, Qiyi Biotechnology, 1:5000) was added and incubated at room temperature for 1 h. After washing again, bands were visualized using the ECL (enhanced chemiluminescence) method. ImageJ software was used for densitometric quantification of the blots, with β-actin used as the reference protein. The ratios of p-PI3K/PI3K, p-Akt/Akt, and MMP-9/β-actin were calculated, and pathway activation differences between the two cell lines were compared.
-
(2)
Gelatin zymography for MMP-9 activity:
Samples were loaded without reducing agent onto a 10% SDS-PAGE (sodium dodecyl sulfate‑polyacrylamide gel electrophoresis) gel containing 0.1% gelatin. Due to the lower protein expression level in C918 cells, the loading amount was adjusted to 40 µg to ensure clear detection signals. All samples were standardized based on total protein amount quantified by the BCA (bicinchoninic acid) method. After electrophoresis, the gel was washed gently with deionized water to remove SDS and then incubated in reaction buffer (50 mM Tris-HCl [pH 7.5], 5 mM CaCl₂, 0.02% NaN₃) at 37 °C for 20 h (shortened to 18 h for MUM-2B cells). The gel was stained with Coomassie Brilliant Blue R-250 for 1 h. After destaining, gelatinolytic bands were quantified using ImageJ software to compare MMP-9 activity between the two cell lines.
-
(3)
PI3K pathway inhibition experiment:
HMGA1-overexpressing (OE) and -knockdown (KD) C918 and MUM-2B cells were treated with 20 µM LY294002 for 24 h (C918 cells for 28 h). Simultaneously, PI3K-siRNA transfection was performed. Cells were collected 48 h later. WB was used to detect the expression of p-PI3K, p-Akt, and MMP-9 to assess the consistency of HMGA1 regulatory effects under pathway inhibition in both cell lines.
Statistical processing
Statistical analysis was performed using SPSS 27.0 software, and graphs were drawn using Origin 2018. All experiments were independently repeated three times. Data were expressed as mean ± standard deviation (Mean ± SD). Comparisons between two groups were performed using Student’s t-test. Comparisons among multiple groups were performed using one-way ANOVA (one‑way analysis of variance), with multiple comparisons conducted using LSD (least significant difference) post-hoc test. The significance level was two-sided, α = 0.05. A P-value < 0.05 was considered statistically significant. Due to software output limitations, only P-value ranges (e.g., P < 0.05, P < 0.01, P < 0.001) could be provided, and exact P-values were unavailable.
Results
Verification of HMGA1 overexpression and knockdown efficiency
To verify the overexpression and knockdown efficiency of HMGA1 in uveal melanoma cells, HMGA1 overexpression plasmid transfection and knockdown using two independent siRNAs (si-1, si-2) were performed in C918 and MUM-2B cells, respectively. Western blot results (Fig. 1) showed that, compared with the control group, HMGA1 protein levels were significantly increased in the overexpression (OE) group. In contrast, HMGA1 protein expression was effectively suppressed in both siRNA-treated knockdown groups (KD1, KD2), with a knockdown efficiency exceeding 70%. qRT‑PCR (quantitative real‑time polymerase chain reaction) results further confirmed that HMGA1 mRNA (messenger RNA) levels were significantly up-regulated in the overexpression group and significantly down-regulated in the knockdown groups, with no significant difference in knockdown efficiency between the two siRNAs.
Fig. 1.
Verification of HMGA1 expression in C918 and MUM-2B cells. Note: OE: Overexpression, KD: Knockdown. A Western blot detection of HMGA1 protein expression levels in C918 and MUM‑2B cells. β‑actin served as the reference protein. B qRT‑PCR detection of relative HMGA1 mRNA expression levels in each group of cells, with GAPDH as the reference gene. Data were presented as the mean ± standard deviation (Mean ± SD) from three independent experiments. Compared with the control group, ***P < 0.001
Effect of HMGA1 on UM cell proliferation
The results (Fig. 2; Tables 1 and 2) showed that the proliferation ability of the HMGA1 overexpression (OE) group was significantly higher than that of the control group in both cell types, while that of the knockdown (KD) group was significantly lower. MTT assay results were normalized to the control group cell viability as 100%. The relative viability of the OE group was 134.6% ± 7.2% in C918 cells and 130.7% ± 6.8% in MUM‑2B cells. EdU assay results showed that the positivity rate in the OE group was 25.3 ± 2.2% (control group: 12.5 ± 1.7%) in C918 cells and 40.5 ± 3.1% (control group: 22.6 ± 2.5%) in MUM‑2B cells. All indicators in the KD groups were significantly reduced, and the differences were statistically significant (P < 0.05).
Fig. 2.
EdU staining fluorescence images (×200). Note: OE: Overexpression, KD: Knockdown. scale bar=200 μm
Table 1.
MTT proliferation results
| Cell line | Group | Day 1 | Day 2 | Day 3 | Day 4 | Statistical result (compared with control group of the same cell line) |
|---|---|---|---|---|---|---|
| C918 | Control | 100.0%±3.1% | 100.0%±3.9% | 100.0%±4.8% | 100.0%±5.0% | - |
| HMGA1-OE | 118.8%±4.2% | 134.1%±5.4% | 134.6%±7.2% | 136.7%±7.8% | P < 0.01 | |
| HMGA1-KD | 78.1%±3.5% | 73.2%±3.2% | 61.5%±4.6% | 63.3%±4.0% | P < 0.001 | |
| MUM-2B | Control | 100.0%±3.6% | 100.0%±4.3% | 100.0%±5.1% | 100.0%±5.6% | - |
| HMGA1-OE | 122.2%±4.5% | 124.1%±5.2% | 130.7%±6.8% | 130.7%±7.3% | P < 0.01 | |
| HMGA1-KD | 66.7%±3.0% | 62.1%±3.4% | 53.3%±4.0% | 51.1%±4.1% | P < 0.001 |
Data were presented as the mean ± standard deviation (Mean ± SD) from three independent experiments
One-way ANOVA was used for intergroup comparison, followed by the LSD post hoc test
Control group cell viability was normalized to 100%
Note: OE Overexpression, KD Knockdown
Table 2.
EdU positive rate results
| Cell line | Group | EdU-Positive rate (%) | Statistical results (vs. control of the same cell line) |
|---|---|---|---|
| C918 | Control | 12.5 ± 1.7 | - |
| HMGA1-OE | 25.3 ± 2.2 | P < 0.01 | |
| HMGA1-KD | 7.8 ± 1.3 | P < 0.001 | |
| MUM-2B | Control | 22.6 ± 2.5 | - |
| HMGA1-OE | 40.5 ± 3.1 | P < 0.01 | |
| HMGA1-KD | 9.2 ± 1.6 | P < 0.001 |
Note: OE Overexpression, KD Knockdown
Effect of HMGA1 on UM cell migration and invasion
The results (Figs. 3 and 4; Table 3) showed that cell migration and invasion capacities were significantly enhanced in the HMGA1-OE groups and significantly weakened in the KD groups. In the scratch healing assay, the 24-hour wound closure rate for C918 OE group was 55.8 ± 4.5% (control: 30.5 ± 4.2%), and for MUM-2B OE group was 78.6 ± 5.3% (control: 50.3 ± 4.8%). In the Transwell assay, the number of invasive cells per field for C918 OE group was 65 ± 7 (control: 32 ± 5), and for MUM-2B OE group was 105 ± 10 (control: 60 ± 8). All indicators in the KD groups were significantly reduced, and all differences were extremely statistically significant (P < 0.001).
Fig. 3.
Scratch healing images (×100). Note: OE: Overexpression, KD: Knockdown. scale bar=200 μm
Fig. 4.
Transwell invasion staining images (×200). Note: OE: Overexpression, KD: Knockdown. scale bar=100 μm
Table 3.
Wound closure rate and invasive cell number results
| Cell line | Group | 24-hour Wound closure rate (%) | Invasive cells per Field | Statistical results (vs. control of the same cell line) |
|---|---|---|---|---|
| C918 | Control | 30.5 ± 4.2 | 32 ± 5 | - |
| HMGA1-OE | 55.8 ± 4.5 | 65 ± 7 | P < 0.001 | |
| HMGA1-KD | 18.2 ± 3.1 | 12 ± 3 | P < 0.001 | |
| MUM-2B | Control | 50.3 ± 4.8 | 60 ± 8 | - |
| HMGA1-OE | 78.6 ± 5.3 | 105 ± 10 | P < 0.001 | |
| HMGA1-KD | 26.5 ± 3.6 | 25 ± 6 | P < 0.001 |
Note: OE Overexpression, KD Knockdown
Regulation of HMGA1 on the PI3K/Akt/MMP-9 pathway
The expression changes of key proteins in the PI3K/Akt pathway and the downstream effector molecule MMP‑9 were examined (Figs. 5 and 6). Western blot results showed that in both C918 and MUM‑2B cells, HMGA1 overexpression (OE) significantly up-regulated the phosphorylation levels of p‑PI3K and p‑Akt, as well as the protein expression of MMP‑9, while having no significant effect on the expression of total PI3K or total Akt. HMGA1 knockdown (KD) significantly suppressed the expression of the aforementioned phosphorylated proteins and MMP‑9. Densitometric quantification results further confirmed that the ratios of p‑PI3K/PI3K, p‑Akt/Akt, and the relative expression level of MMP‑9 were significantly higher in the overexpression group compared with the control group, and significantly lower in the knockdown group compared with the control group, with all differences being statistically significant (P < 0.01; P < 0.001).
Fig. 5.
Expression of key pathway proteins. Note: OE: Overexpression, KD: Knockdown. A Western blot detection of p‑PI3K, PI3K, p‑Akt, Akt, and MMP‑9 protein expression levels in C918 and MUM‑2B cells. β‑actin served as the reference protein. B Densitometric quantification of p‑PI3K/PI3K, p‑Akt/Akt, and MMP‑9/β‑actin ratios in each group of cells. Data were presented as the mean ± standard deviation (Mean ± SD) from three independent experiments. Compared with the control group, **P < 0.01, ***P < 0.001
Fig. 6.
Gelatin zymography for MMP-9 activity. Note: OE: Overexpression, KD: Knockdown. A Gelatin zymography detection of MMP‑9 enzymatic activity in C918 and MUM-2B cells. B Relative quantitative analysis of MMP‑9 enzymatic activity. Data were presented as the mean ± standard deviation (Mean ± SD) from three independent experiments. Compared with the control group, ***P < 0.001
Reversal effect of LY294002 on HMGA1 effects
The results in Fig. 7 show that after treatment with 20 µM LY294002, the activation level of the PI3K/Akt pathway in HMGA1-OE cells was significantly reduced, and their proliferation and migration capacities were also significantly reversed. In C918 cells after treatment, p-PI3K/PI3K was 0.85 ± 0.09 (OE group: 1.50 ± 0.10), MTT OD570 value was 0.53 ± 0.05 (OE group: 0.70 ± 0.06), and the 24-hour wound closure rate was 35.2 ± 4.1% (OE group: 55.8 ± 4.5%). In MUM-2B cells after treatment, p-PI3K/PI3K was 1.12 ± 0.11 (OE group: 2.05 ± 0.15), MTT OD570 value was 0.75 ± 0.06 (OE group: 0.98 ± 0.07), and the 24-hour wound closure rate was 48.3 ± 4.6% (OE group: 78.6 ± 5.3%). All differences were statistically significant (P < 0.05, P < 0.01).
Fig. 7.
Pathway protein phosphorylation levels in OE group and OE + inhibitor group. Note: OE: Overexpression, KD: Knockdown. A Western blot detection of p‑PI3K, PI3K, p‑Akt, Akt, and MMP‑9 protein expression in C918 and MUM‑2B cells, with β‑actin as the reference B Densitometric quantification of p‑PI3K/PI3K, p‑Akt/Akt, and MMP‑9/β‑actin ratios. Data were presented as the mean ± standard deviation from three independent experiments. Compared with the Control group: ##P < 0.01, ###P < 0.001; compared with the OE group: *P < 0.05, **P < 0.01, ***P < 0.001
Discussion
By modulating HMGA1 expression in uveal melanoma cell lines from different sources, this study systematically investigated its promoting effect on cell proliferation, migration, and invasion. The results confirmed that HMGA1 overexpression significantly up-regulated HMGA1 mRNA and protein levels in all cell lines, whereas knockdown significantly reduced its expression, suggesting that HMGA1 plays a consistently malignant regulatory role in different UM cells. Previous studies have shown that HMGA1 is highly expressed and drives tumor progression in various tumors, including liver cancer and gastric cancer [19, 20]. The above results confirmed that the transfection model was effectively established and that UM cells are generally sensitive to changes in HMGA1 expression, whose level can directly affect the malignant phenotype of the cells, laying the foundation for subsequent functional experiments. It is speculated that this general sensitivity may stem from the conserved function of HMGA1 as a key regulatory factor in UM cells, with high dependency on HMGA1 regardless of whether the cells are from primary or metastatic origins.
Regarding proliferation ability, HMGA1 overexpression significantly enhanced the proliferative activity of UM cells, while knockdown significantly suppressed proliferation. This trend was consistent with the role of HMGA1 in regulating proliferation in lung cancer and also aligned with the regulatory patterns of other pro-cancer molecules in UM [21]. Notably, the highly metastatic MUM‑2B cells responded more significantly to HMGA1 regulation, suggesting that the pro-proliferative effect of HMGA1 may be closely related to the metastatic potential of UM, which is consistent with the role of HMGA1 in metastatic tumors [22]. It is speculated that during tumor progression, as metastatic ability increases, cells become more dependent on HMGA1-mediated proliferation signals. Mechanistically, HMGA1 overexpression significantly activated the PI3K/Akt pathway, which was consistent with reports of HMGA1 regulating this pathway in colorectal cancer [23, 24], suggesting that this regulatory pattern is conserved across different tumors. HMGA1 may promote Akt phosphorylation by enhancing PI3K activity, thereby regulating cyclins and apoptosis-related proteins, ultimately driving cell proliferation.
In migration and invasion assays, HMGA1 overexpression significantly enhanced the migration and invasion abilities of UM cells, while knockdown significantly suppressed these processes. This result was consistent with the pro-metastatic role of HMGA1 in gastric cancer and further confirmed the universality of its pro-cancer function. This suggested that HMGA1 not only enhances cell motility but also promotes basement membrane invasion, with a more pronounced effect in highly metastatic cells, which aligns with the molecular characteristics of UM metastasis [25]. Concurrently, HMGA1 significantly up-regulated MMP‑9 expression, while its down-regulation reduced MMP‑9 levels, suggesting that MMP‑9 is a key downstream molecule through which HMGA1 regulates UM invasion. This finding is consistent with reports on the central role of MMP‑9 in UM invasion and metastasis [26]. It also suggested that HMGA1 may up-regulate MMP‑9 through transcriptional activation or chromatin remodeling mechanisms. As a key enzyme that degrades collagen and the basement membrane, increased MMP‑9 activity reduces the matrix barrier, creating conditions for UM cell metastasis, which aligns with previous findings on UM metastasis mechanisms [27].
Furthermore, following intervention with the PI3K inhibitor LY294002, the enhanced proliferation and migration induced by HMGA1 overexpression were significantly reversed to near-control levels. This rescue effect validated the specificity of the pathway regulation and was consistent with the effects of PI3K inhibitors in other tumors. Additionally, this rescue experiment confirmed that PI3K/Akt is a key pathway mediating the biological functions of HMGA1. As an upstream molecule, HMGA1 relies on PI3K/Akt activation and regulates MMP‑9 transcription through downstream transcription factors [28]. In summary, HMGA1 expression levels were positively correlated with PI3K/Akt activity and MMP‑9 levels, with consistent effects across UM cells from different sources. The magnitude of the effect correlated with the degree of malignancy of the cells, suggesting that HMGA1 may serve as a potential marker of UM malignancy. These findings not only establish the core role of HMGA1 in UM progression but also validate the signal amplification function of the PI3K/Akt/MMP-9 axis [29], suggesting that therapeutic strategies targeting HMGA1 or its downstream pathway may exert inhibitory effects on UM variants with different invasiveness, providing an experimental basis for the development of universal molecular targeted interventions.
This study has certain limitations. It only validated the core mechanism based on in vitro cell experiments, without further confirmation through in vivo animal models or large-scale clinical samples. The upstream regulatory mechanisms of HMGA1 and its role in the tumor microenvironment were not deeply explored. In the future, UM animal models could be constructed to verify the effects of HMGA1 on tumor growth and invasion in vivo. Simultaneously, clinical UM patient samples could be collected to analyze the correlation between HMGA1 expression levels and patient prognosis, further clarifying its clinical significance. Additionally, the interaction between HMGA1 and UM-specific driver genes could be explored in depth to define its position within the UM molecular network, providing a basis for developing more precise combination treatment strategies.
Conclusion
This study confirmed through in vitro cell experiments that HMGA1 significantly promotes the proliferation, migration, and invasion capacities of UM cells from different origins by activating the PI3K/Akt signaling pathway and upregulating MMP-9 expression. This finding clarifies the core role and molecular mechanism of HMGA1 in the malignant progression of UM, suggesting that HMGA1 can serve as a potential target for molecular targeted therapy of UM. It also provides experimental evidence for the development of prognostic biomarkers for UM. This study was conducted solely based on in vitro experiments and has certain limitations. Future validation through in vivo animal models and clinical samples is needed to further elucidate the clinical significance of HMGA1, providing more substantial theoretical and experimental support for the development of precise treatment strategies for UM.
Supplementary Information
Acknowledgements
Not applicable.
Authors’ contributions
WR is responsible for the study design, overall supervision, and drafting the manuscript; BL constructed the HMGA1 overexpression and knockdown cell models and performed the functional experiments; WR completed the Western blot and gelatin zymography analyses and conducted the data statistics. All authors contributed to the discussion of the results and reviewed and approved the final manuscript.
Funding
Not applicable.
Data availability
The datasets analyzed in the current study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Carvajal RD, Sacco JJ, Jager MJ, Eschelman DJ, Olofsson Bagge R, Harbour JW, et al. Piperno-Neumann S. Advances in the clinical management of uveal melanoma. Nat Reviews Clin Oncol. 2023;20(2):99–115. [DOI] [PubMed] [Google Scholar]
- 2.Heiferman MJ, Mahajan VB, Mruthyunjaya P. Proteomics in uveal melanoma. Curr Opin Ophthalmol. 2022;33(3):202–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Rantala ES, Hernberg MM, Piperno-Neumann S, Grossniklaus HE, Kivelä TT. Metastatic uveal melanoma: The final frontier. Prog Retin Eye Res. 2022;90:101041. [DOI] [PubMed] [Google Scholar]
- 4.Bustamante P, Piquet L, Landreville S, Burnier JV. Uveal melanoma pathobiology: Metastasis to the liver. Sem Cancer Biol. 2021;71:65–85. [DOI] [PubMed] [Google Scholar]
- 5.Liu H, Tang T. Pan-cancer genetic analysis of disulfidptosis-related gene set. Cancer Genet. 2023;278–279:91–103. [DOI] [PubMed] [Google Scholar]
- 6.Chang X, Liu J, Yang Q, Gao Y, Ding X, Zhao J, et al. Targeting HMGA1 contributes to immunotherapy in aggressive breast cancer while suppressing EMT. Biochem Pharmacol. 2023;212:115582. [DOI] [PubMed] [Google Scholar]
- 7.Sun Y, Guo G, Zhang Y, Chen X, Lu Y, Hong R, et al. IKBKE promotes the ZEB2-mediated EMT process by phosphorylating HMGA1a in glioblastoma. Cell Signal. 2024;116:111062. [DOI] [PubMed] [Google Scholar]
- 8.Li Z, Liu J, Chen T, Sun R, Liu Z, Qiu B, et al. HMGA1-TRIP13 axis promotes stemness and epithelial mesenchymal transition of perihilar cholangiocarcinoma in a positive feedback loop dependent on c-Myc. J Experimental Clin Cancer Res. 2021;40(1):86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Liu H. Effect of traditional medicine on clinical cancer. Biomedical J Sci Tech Research. 2020;30.
- 10.He J, Zeng F, Jin XI, Liang L, Gao M, Li W, Li G, Zhou Y. YWHAH activates the HMGA1/PI3K/AKT/mTOR signaling pathway by positively regulating Fra-1 to affect the proliferation of gastric cancer cells. Oncol Res. 2023;31(4):615–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhang BC, Ma SY, Zhu P, Zhu LY, Zhao XX, Pu C. LINC00665 target let-7i/HMGA1 promotes the proliferation and invasion of hepatoma cells. Mutat Res. 2024;828:111852. [DOI] [PubMed] [Google Scholar]
- 12.Tian Z, Liu Z, Fang X, Cao K, Zhang B, Wu R, et al. ANP32A promotes the proliferation, migration and invasion of hepatocellular carcinoma by modulating the HMGA1/STAT3 pathway. Carcinogenesis. 2021;42(3):493–506. [DOI] [PubMed]
- 13.Bai Y, Liu F, Yang Z. CircRNA LRP6 promotes high-glucose induced proliferation and migration of vascular smooth muscle cells through regulating miR-545-3p/HMGA1 signaling axis. Am J Translational Res. 2021;13(8):8909–20. [PMC free article] [PubMed] [Google Scholar]
- 14.Qu J, Li J, Zhang Y, He R, Liu X, Gong K, et al. AKR1B10 promotes breast cancer cell proliferation and migration via the PI3K/AKT/NF-κB signaling pathway. Cell Bioscience. 2021;11(1):163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang J, Yang K, Yang X, Jin T, Tian Y, Dai C, et al. HHLA2 promotes hepatoma cell proliferation, migration, and invasion via SPP1/PI3K/AKT signaling pathway. Mol Carcinog. 2024;63(7):1275–87. [DOI] [PubMed] [Google Scholar]
- 16.Wang Q, Wang B, Zhang W, Zhang T, Liu Q, Jiao X, et al. APLN promotes the proliferation, migration, and glycolysis of cervical cancer through the PI3K/AKT/mTOR pathway. Arch Biochem Biophys. 2024;755:109983. [DOI] [PubMed] [Google Scholar]
- 17.Chen X, Wu Q, Depeille P, Chen P, Thornton S, Kalirai H, Coupland SE, Roose JP, Bastian BC. RasGRP3 Mediates MAPK Pathway Activation in GNAQ Mutant Uveal Melanoma. Cancer Cell. 2017;31(5):685–e6966. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Salzano S, Caltabiano R, Zanelli M, Palicelli A, Zizzo M, Boutas I, Koufopoulos N, Broggi G. Preferentially Expressed Antigen in Melanoma (PRAME) as a diagnostic and predictive marker in melanocytic tumors: An updated narrative review. Histol Histopathol. 2025;40(12):1889–95. [DOI] [PubMed] [Google Scholar]
- 19.Zhu J, Zheng Y, Liu Y, Chen M, Liu Y, Li J. Association between HMGA1 and immunosuppression in hepatocellular carcinoma: A comprehensive bioinformatics analysis. Med (Baltim). 2023;102(4):e32707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ouyang J, Li J, Li D, Jiang J, Hao T, Xia Y, Lu X, Zhang C, He Y. IGF2BP2 Promotes Epithelial to Mesenchymal Transition and Metastasis through Stabilizing HMGA1 mRNA in Gastric Cancer. Cancers (Basel). 2022;14(21):5381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ma Y, Ma X, Du B, Li X, Li Y. HMGA1 is a Prognostic Biomarker and Correlated with Glycolysis in Lung Adenocarcinoma. J Cancer. 2024;15(10):2913–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Gao C, Xie Q, Li L, Luo S, Fan J, Cao J. HMGA2/PHGDH axis-mediated glycolysis under hypoxia is required for cadmium-induced A549 cell migration. Ecotoxicol Environ Saf. 2025;303:118787. [DOI] [PubMed] [Google Scholar]
- 23.Wang L, Zhang J, Xia M, Liu C, Zu X, Zhong J. High Mobility Group A1 (HMGA1): Structure, Biological Function, and Therapeutic Potential. Int J Biol Sci. 2022;18(11):4414–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Sonkin D, Thomas A, Teicher BA. Cancer treatments: Past, present, and future. Cancer Genet. 2024;286–287:18–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.He KY, Zhao A, Guo JR, Wu DH, Liu H, Gao F, Liu MJ, Yang JY, Lei XY, Li JQ, Zhang L, Yan ZH, Ding Q, Huang YW, Cui R, Jian YP, Xu ZX. High mobility group A1 (HMGA1) promotes esophageal squamous cell carcinoma progression by inhibiting STING-mediated anti-tumor immunity. Nat Commun. 2025;16(1):5098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Salatino A, Mirabelli M, Chiefari E, Greco M, Di Vito A, Bonapace G, Brunetti FS, Crocerossa F, Epstein AL, Foti DP, Brunetti A. The anticancer effects of Metformin in the male germ tumor SEM-1 cell line are mediated by HMGA1. Front Endocrinol (Lausanne). 2022;13:1051988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Labarrade F, Garcia I, Imbert I. Collagen type V alpha 3 chain is involved in human skin basement membrane physiology and MMP-9 regulation. Int J Cosmet Sci. 2025;47(5):743–51. [DOI] [PubMed] [Google Scholar]
- 28.Akade E, Jalilian S. The role of high mobility group AT-hook 1 in viral infections: Implications for cancer pathogenesis. Int J Biochem Cell Biol. 2024;169:106532. [DOI] [PubMed] [Google Scholar]
- 29.Gu H, Wang C, Li J, Yang Y, Sun W, Jiang C, et al. High mobility group box-1-toll-like receptor 4-phosphatidylinositol 3-kinase/protein kinase B-mediated generation of matrix metalloproteinase-9 in the dorsal root ganglion promotes chemotherapy-induced peripheral neuropathy. Int J Cancer. 2020;146(10):2810–21. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets analyzed in the current study are available from the corresponding author upon reasonable request.