GRP75-upregulated HMGA1 Expression Promotes Stage I Lung Adenocarcinoma Progression by Activating the JNK/c-JUN Signaling Pathway

Background: Recurrence is a major challenge in early-stage lung adenocarcinoma (LUAD) treatment. However, the recurrence mechanism is still unclear, and no biomarkers can predict recurrence in early-stage LUAD. Here, we investigated the role and mechanism of high-mobility group AT-hook 1 (HMGA1) and glucose-regulated protein 75-kDa (GRP75) in stage I LUAD and evaluated their potential as biomarkers for predicting the recurrence and prognosis of stage I LUAD. Methods: A TCGA dataset was used to investigate the clinical signicance of HMGA1 and GRP75 in early-stage LUAD. Western blotting and immunohistochemistry were used to measure protein expression levels. The biological functions of HMGA1 and GRP75 in LUAD were investigated both in vitro and in vivo through overexpression and knockdown experiments. The interaction and regulation between HMGA1 and GRP75 were evaluated with coimmunoprecipitation and ubiquitination assays. The downstream signaling pathway of the GRP75/HMGA1 axis was investigated by mRNA-sequencing analysis. Results: High expression of HMGA1 and GRP75 was associated with recurrence and a poor prognosis in stage I LUAD patients. In particular, HMGA1 had potential as an independent prognostic factor. Overexpression of GRP75 or HMGA1 promoted LUAD cell growth and metastasis, while silencing GRP75 or HMGA1 inhibited LUAD cell growth and metastasis. In vitro and clinical data showed that the expression level of GRP75 positively regulated HMGA1 in LUAD and that GRP75 played an HMGA1-dependent role. In addition, GRP75 prolonged the half-life of HMGA1 by inhibiting HMGA1 ubiquitination via direct binding to HMGA1. Finally, we demonstrated that the GRP75/HMGA1 axis played a role by activating JNK/c-JUN signaling in LUAD. Conclusions: PTM:Post-translational modication; JNKs:c-JUN N-terminal chaperone mortalin contributes to epithelial-mesenchymal transition metastasis.


Background
Lung cancer is the most malignant human cancer with the highest incidence and mortality worldwide [1].
Lung adenocarcinoma (LUAD) is a type of non-small cell lung cancer (NSCLC) that accounts for 40% of all lung malignancies [2] and surgery is the major treatment method for patients with early-stage LUAD [3,4]. However, 30%~40% of stage I LUAD patients ultimately die because of locoregional or metastatic recurrence after surgery [5]. Unfortunately, no biomarker can accurately predict postoperative recurrence in patients with early-stage LUAD, and the postoperative progression mechanism of early-stage LUAD is currently unclear.
High-mobility group AT-hook 1 (HMGA1) is an architectural transcription factor [6]. Studies show that HMGA1 expression is abnormally upregulated in several types of cancer, including lung cancer, and HMGA1 acts as an oncogene [7][8][9]. Notably, various studies indicate that HMGA1 is involved in the regulation of epithelial-mesenchymal transition (EMT) and cancer stem cells (CSCs) [10][11][12]. EMT is a key step in cancer metastasis because it is a process by which epithelial tumor cells lose their cell polarity and cell-cell adhesion and gain migratory and invasive properties [13]. Additionally, a study showed that recurrence is accompanied by EMT in breast cancer [14]. Additionally, accumulating evidence shows that CSCs are another key factor in cancer metastasis and recurrence [15][16][17]. CSCs are a small subset of cells within tumors that behave like stem cells. Studies show that CSCs have high tumorigenicity, are able to undergo self-renewal, and can regenerate all cell types in a tumor, resulting in tumor relapse [18].
Importantly, clinical data also show that the expression levels of the CSC marker protein ALDH1 and CD133 are closely correlated with recurrence and a poor prognosis in patients with early-stage NSCLC [19]. Together, these ndings suggest that HMGA1 may be involved in the progression of early-stage lung cancer. However, the upregulation mechanism and the functional role of HMGA1 in the progression of early stage LUAD have not been reported.
In the present study, we demonstrated that HMGA1 stimulated LUAD cell viability and metastasis and a high expression level of HMGA1 was an independent prognostic indicator closely correlated with recurrence and a poor prognosis in patients with stage I LUAD. In addition, we found that the upregulated expression of HMGA1 was caused by increased expression of GRP75 in LUAD. Additionally, we identi ed that GRP75 was associated with recurrence and a poor prognosis in stage I LUAD and played an oncogenic role that was HMGA1 dependent. Finally, we demonstrated that the GRP75/HMGA1 axis promoted LUAD progression by activating JNK/c-JUN signaling.

Materials
Fetal bovine serum (FBS) and cell culture medium were purchased from HyClone (South Logan, UT, USA). MG132 and kanamycin were purchased from Sigma (St. Louis, MO, USA). JNKIN8 was obtained from MedChem Express (Shanghai, China). Plasmids for overexpressing HMGA1 or GRP75 and HMGA1speci c short hairpin RNA (shRNA) were designed and synthesized by GeneChem (Shanghai, China).

Cell lines, cell transfection and RNA interference
The lung adenocarcinoma cell lines A549 and PC9 were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA). All cell lines were maintained in Dulbecco's modi ed Eagle's medium supplemented with 10% FBS at 37 °C in a humidi ed atmosphere with 5% CO 2 .

Transwell and cell viability assays
The migratory and invasive abilities of LUAD cells were assessed with a transwell assay. The indicated cells were transfected with the indicated oligonucleotides or plasmids for 48 h. Then, 3 × 10 4 cells in serum-free growth medium were seeded in the upper chambers. The lower chambers were lled with the same medium supplemented with 10% FBS. After 24 h, the cells that had invaded the lower side of the chamber were xed with 4% paraformaldehyde (PFA), stained with 0.1% crystal violet and dried. The number of invading cells was counted under a light microscope.
For cell viability assays, the indicated cells were plated in 96-well plates at 4 × 10 3 cells per well. At the indicated time points, cell viability was estimated by using CCK8 reagent according to the manufacturer's instructions. Absorbance was measured at a wavelength of 450 nm by using a microplate reader.

Colony formation assays
After 48 h of transfection, cells were trypsinized and resuspended in 0.5 ml 0.35% agar in growth medium at a density of 1000 cells/well (24-well plate). Then, the agar-cell mixture was plated on top of a solid layer of 0.8% agar in the growth medium. The cells were cultured until visible clones appeared (approximately 3 weeks). The cell clones were washed twice with PBS, xed with methanol for 20 min, stained with 0.1% crystal violet for 1 h, washed with PBS, and counted.

Western blot analysis
Western blot analysis was carried out according to a standard protocol. Cell lysates were generated from cultured cells with RIPA lysis buffer containing protease inhibitors (Roche). The protein concentration was determined by using a Bradford kit (Pierce, Thermo Fisher Scienti c). Equal amounts of total protein were loaded, separated by SDS-PAGE and then transferred to PVDF membranes (Bio-Rad, Hercules, CA, USA).
After blocking in 5% nonfat milk, the membranes were incubated overnight at 4 °C with appropriate primary antibodies and then with secondary antibodies conjugated to horseradish peroxidase for 2 hours at room temperature. After washing, the bands of interest were analyzed by using an ECL kit. Images were acquired using the Bio-Spectrum Gel Imaging System (Bio-Rad, USA). Then, quanti cation of the Western blot results was performed using Image Lab software, and the bands were normalized with GAPDH as the internal control.
Immunoprecipitation and mass spectrometry Immunoprecipitation was performed using a Thermo Fisher Scienti c Pierce coimmunoprecipitation kit following the manufacturer's protocol. Brie y, the indicated antibodies were rst immobilized for 2 h using AminoLink Plus coupling resin. The resin was then washed and incubated with lysate overnight at 4 °C. After the incubation, the resin was washed, and proteins were eluted using elution buffer. A negative control that was provided with the co-IP kit to assess nonspeci c binding received the same treatment as the samples with IgG. The immunoprecipitated samples were resolved by SDS-PAGE, followed by Western blotting with appropriate antibodies.
For liquid chromatography-mass spectrometry (LC-MS) analysis, samples, which included those with HMGA1 overexpression and a control, were separated using electrophoresis, and speci c bands were identi ed using a mass spectrometry system; the molecules corresponding to the bands were identi ed in a human proteomic library to obtain proteins directly bound to HMGA1.

Degradation and deubiquitylation assay
To detect HMGA1 degradation mediated by GRP75 in A549 cells, cells with or without GRP75 overexpression were treated with CHX (CST, 100 µg/ml) for the indicated durations before analysis. Then, cell lysates were analyzed using Western blotting with an anti-HMGA1 antibody. To detect HMGA1 deubiquitination in A549 cells, cells were transfected with the indicated plasmids and treated with 20 µM proteasome inhibitor MG132 (Sigma) for 8 h before harvest. Then, the cells were lysed in IP lysis/wash buffer and incubated with the anti-HMGA1 antibody, which was used for IP. The immunoprecipitated samples were resolved by SDS-PAGE, followed by Western blotting with an anti-ubiquitin antibody.

Bioinformatic analysis
Publicly available databases were used to evaluate gene expression and predict potential mechanisms. A clinical dataset of stage I LUAD patients was obtained from the TCGA database and analyzed using TIMER [20]. mRNA-sequencing analysis was performed by Huada Genomics Institute (BGI, Guangzhou, China), and KEGG pathway analysis was conducted with the Dr. Tom program (https://biosys.bgi.com).

Animal experiments
Animal studies were conducted at Daping Hospital, Army Medical University, according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All animal experiments were approved by the Animal Research Committee of Army Medical University. Six-week-old female nude mice were utilized. To investigate the effects of oncogenic HMGA1 on LUAD cell tumorigenicity, the indicated cells were transfected with HMGA1 or HMGA1-speci c shRNA-expressing or control plasmids. After 72 h of transfection, 2 × 10 6 cells in 0.1 mL PBS were injected subcutaneously ( ve mice per group). Tumor length (L) and width (W) were measured weekly, and tumor volume (V) was calculated using the formula V = L × W 2 /2. Four weeks after cell injection, these mice were sacri ced, and the tumors were harvested and weighed. To investigate the presence of Ki67 proteins in mouse tumor tissues, an immunohistochemistry assay was performed as described previously [21].

Statistical analysis
All data are presented as the mean ± SD, and statistical analyses were performed using GraphPad Prism 8 (GraphPad Software, Inc., La Jo lla, CA, USA). The statistical signi cance of differences between treatment groups was analyzed by a two-tailed paired t-test or one-way analysis of variance (ANOVA). The survival rate was calculated using Kaplan-Meier survival analysis. P < 0.05 was considered statistically signi cant.

Results
Upregulated expression of HMGA1 is associated with recurrence and a poor prognosis in stage I LUAD patients First, we used a TCGA dataset to investigate whether the expression level of HMGA1 is involved in the recurrence of stage I LUAD. As in previous reports, the TCGA analysis results showed that HMGA1 expression was abnormally upregulated in many types of cancer tissue, including LUAD tissue, compared to the corresponding normal tissue ( Fig. 1a and b). Notably, the expression level of HMGA1 in stage I LUAD tissues was signi cantly higher than that in normal lung tissues (Fig. 1c), and the high expression level of HMGA1 in primary tumors was closely correlated with recurrence in patients with stage I LUAD (Fig. 1d). In addition, the TCGA dataset analysis results showed that the stage I LUAD patient group with a high level of HMGA1 had a shorter median recurrence-free survival (RFS) time (Fig. 1e) and lower overall survival (OS) rate than the low HMGA1 expression group (Fig. 1f). Collectively, these results indicate that the increased expression level of HMGA1 is closely correlated with recurrence and a poor prognosis in stage I LUAD patients.
Upregulation of HMGA1 expression stimulates LUAD progression and is a therapeutic target in LUAD Next, we investigated whether the upregulated expression of HMGA1 directly stimulates LUAD progression. We rst generated a construct expressing green uorescent protein (GFP)-tagged HMGA1 and transfected it into LUAD cells (Fig. 2a), we then performed cell viability, soft agar colony formation and transwell assays. Our results showed that overexpression of HMGA1 signi cantly stimulated cell viability (Fig. 2b), soft agar colony formation (Fig. 2c), migration and invasion (Fig. 2d) in both A549 cells and PC9 cells. These in vitro results were further con rmed in vivo using xenograft models. As shown in Fig. 2e and 2f, tumor growth and tumor weight were remarkably higher in the HMGA1 overexpression group than in the vector control group. Additionally, an increased expression level of the cell proliferation marker protein Ki67 was detected in HMGA1-overexpressing LUAD tissues (Fig. 2g). These ndings suggest that upregulated expression of HMGA1 directly stimulate LUAD progression.

Upregulated protein expression of HMGA1 was caused by increased expression of GRP75 in LUAD
To elucidate the regulatory and functional mechanism of HMGA1 overexpression in LUAD, we identi ed proteins that bind to HMGA1 in LUAD cells using IP/LC-MS assays. Our results identi ed 25 proteins that might speci cally interact with HMGA1 in LUAD cells (Figs. 4a and b, Table S1). Among them, GRP75 was selected for further study, as GRP75 was previously reported to promote metastasis and be closely associated with early recurrence in liver cancer [22]. The interaction between GRP75 and HMGA1 in LUAD was further con rmed using coimmunoprecipitation (co-IP) assays (Fig. 4c, Fig. S2). Notably, TCGA data analyses showed that GRP75 protein levels were positively correlated with HMGA1 protein levels in stage I LUAD tissues (Fig. 4d). Since the role of GRP75 in early-stage LUAD is not clear, we then investigated the effect of GRP75 on early-stage LUAD using a TCGA dataset analysis and in vitro experiments. As shown in Fig. 4e-g, the TCGA dataset analysis showed that GRP75 was abnormally highly expressed in stage I LUAD tissues compared to normal tissues (Fig. 4e) and that high expression levels of GRP75 were closely associated with recurrence (Fig. 4f) and shortened RFS in patients with stage I LUAD (Fig. 4g). In addition, the in vitro experiments showed that overexpression of GRP75 signi cantly stimulated LUAD cell viability (Fig. 4h), migration and invasion (Fig. 4i), while silencing GRP75 signi cantly suppressed LUAD cell viability (Fig. 4j), migration and invasion (Fig. 4k). These ndings suggest that GRP75 acts as an oncogene in early-stage LUAD and may be a partner protein of HMGA1.
Next, we investigated how HMGA1 and GRP75 affect each other's expression and functions in LUAD cells. Our Western blot data showed that HMGA1 overexpression or silencing did not affect GRP75 protein levels, while GRP75 overexpression or silencing upregulated or downregulated HMGA1 protein levels in LUAD cell lines, respectively (Fig. 5a). In addition, in vitro experiments showed that silencing of HMGA1 signi cantly inhibited the GRP75 overexpression-induced stimulation of LUAD cell viability (Fig. 5b), invasion and migration (Fig. 5c). Interestingly, multivariate analysis identi ed HMGA1 but not GRP75 as an independent prognostic predictor in stage I LUAD (Table 1). Importantly, RFS analysis showed that there was no correlation between a high expression level of GRP75 and recurrence in stage I LUAD patients with low expression of HMGA1 (Fig. 5d). Taken together, these ndings suggest that HMGA1 is a downstream protein of GRP75 and that GRP75 promotes HMGA1-dependent early-stage LUAD progression. Further, we investigated the mechanism by which GRP75 upregulates HMGA1 expression in LUAD. Unlike our protein data, our mRNA data showed that the mRNA expression level of HMGA1 was not affected by GRP75 overexpression or silencing (Fig. 5e), suggesting that GRP75 regulates HMGA1 expression through posttranscriptional regulation. Because GRP75 binds directly to HMGA1, we speculated that GRP75 may participate in the regulation of HMGA1 protein degradation. To con rm this hypothesis, we investigated the effect of GRP75 on the half-life of the HMGA1 protein in LUAD cells treated with the protein synthesis inhibitor cycloheximide (CHX). Compared with the vector control, GRP75 overexpression robustly prolonged the half-life of the endogenous HMGA1 protein in A549 cells (Fig. 5f). In addition, the ubiquitination of HMGA1 in A549 cells was dramatically inhibited by GRP75 overexpression (Fig. 5g). These ndings suggest that GRP75 upregulates the HMGA1 protein level by protecting the HMGA1 protein from ubiquitination-mediated degradation.
GRP75/HMGA1 axis stimulates LUAD progression by activating the JNK/c-JUN signaling pathway To further investigate the molecular mechanisms by which the GRP75/HMGA1 axis stimulates the progression of early-stage LUAD, we performed mRNA sequencing using HMGA1-or GRP75overexpressing LUAD cells and their corresponding control cells (Fig. 6a). Then, KEGG signaling pathway enrichment analysis was performed using a set of genes that similarly upregulated or downregulated in cells with overexpression of HMGA1 and GRP75 compared to their corresponding control. As shown in Fig. 6b, we found that the JNK signaling pathway was correlated with overexpressed HMGA1 and GRP75 in LUAD. In addition, our Western blot analysis showed that overexpression of HMGA1 stimulated the phosphorylation of JNK (Thr183/Tyr185) and its downstream protein c-JUN (Ser63 and Ser73) in LUAD cells, whereas HMGA1 silencing inhibited the phosphorylation of both JNK (Thr183/Tyr185) and c-JUN (Ser63 and Ser73) (Fig. 6c). Consistent with the HMGA1 results, overexpression of GRP75 also increased the phosphorylation of both JNK and c-JUN in LUAD cells, and the upregulation of phosphor-JNK and phosphor-c-JUN levels induced by GRP75 overexpression was blocked by silencing HMGA1 (Fig. 6d).

Discussion
Here, we report for the rst time that upregulated expression levels of both GRP75 and HMGA1 are closely correlated with recurrence and a poor prognosis in stage I LUAD patients. In particular, HMGA1 has great potential as an independent biomarker for predicting the recurrence and poor prognosis of stage I LUAD patients. Our results are supported by research from other groups. Although not in LUAD, similar results have been obtained in other tumors. According to Yi et al., a high expression level of GRP75 is closely correlated with liver cancer early recurrence [22]. According to Liu et al., high expression of HMGA1 is closely correlated with glioblastoma recurrence [23]. However, the sample size used in this study is limited, so the nding needs to be further veri ed in a larger clinical sample before clinical use.
Here, we also clari ed the mechanism of HMGA1 upregulation in LUAD. GRP75 is an oncogene that is overexpressed in several cancers [24] and involved in posttranslational modi cation of some transcription factors, especially ubiquitinylation [25]. For instance, GRP75 stabilizes the proproliferative association of Hsp90 and EGFR by rescuing the Tid1L-promoted ubiquitinylation and degradation of EGFR, thereby contributing to counteracting death pathways [26]. In this study, we identi ed that GRP75 was abnormally upregulated in stage I LUAD and that overexpression of GRP75 promoted LUAD cell progression. In addition, our in vitro data showed that overexpression of GRP75 upregulated HMGA1 protein levels, GRP75 directly bound to HMGA1, GRP75 prolonged the half-life of the HMGA1 protein and GRP75 inhibited HMGA1 ubiquitination. Notably, the expression level of GRP75 was positively correlated with the HMGA1 expression level in LUAD. Together, these ndings suggest that abnormally overexpressed GRP75 contributes to early-stage LUAD progression and causes HMGA1 upregulation in LUAD by inhibiting ubiquitination-mediated HMGA1 degradation via direct binding to HMGA1. However, the detailed mechanism by which GRP75 inhibits HMGA1 ubiquitination and the mechanism underlying GRP75 upregulation in stage I LUAD need further study.
Finally, we elucidated the oncogenic mechanism of GRP75 and HMGA1 in LUAD progression. Clari cation of the molecular mechanisms of tumor occurrence and progression is urgently needed to help identify effective therapeutic targets. The oncogenic role and therapeutic potential of GRP75 or HMGA1 have been reported in several cancers [27][28][29]. According to Resar et al., HMGA1 appears to drive transformation in human lung cancer [30], and inhibition of HMGA1 can dramatically inhibit lung cancer cell growth and migration [31,32]. Additionally, several studies have indicated that GRP75 stimulates cell growth, proliferation and stemness [33][34][35] and that inhibiting GRP75 can signi cantly inhibit cancer progression [36]. Consistently, the present study results also show that both GRP75 and HMGA1 play oncogenic roles in LUAD and exhibit potential as therapeutic targets in LUAD. However, the highlight of this study is that we demonstrated for the rst time that GRP75 depends on HMGA1 to play an oncogenic role in LUAD. Notably, we identi ed that the GRP75/HMGA1 axis plays an oncogenic role by activating JNK/c-JUN signaling. Various ndings have reported that activation of the JNK/c-JUN signaling pathway is closely correlated with cancer progression, including recurrence [37][38][39]. For instance, Jorgense et al.
reported that activation of c-JUN was associated with cell proliferation and a shortened relapse-free period in super cially spreading malignant melanoma [40]. According to Hagiwara et al., activation of c-JUN predicts a high risk of recurrence after hepatic resection for hepatocellular carcinoma [41]. To the best of our knowledge, this is the rst evidence that HMGA1, as a critical target gene of GRP75, activates the JNK/c-JUN signaling pathway in the progression mechanism of lung adenocarcinoma. However, the detailed mechanism by which HMGA1 activates the JNK pathway in LUAD requires further study.

Conclusions
Our study combined TCGA dataset analysis and experimental studies to establish the oncogenic roles of HMGA1 and GRP75 in stage I LUAD and the mechanism of HMGA1 upregulation. We show that GRP75 causes upregulation of HMGA1 expression by inhibiting ubiquitination through binding to HMGA1 in LUAD and that upregulated HMGA1 expression promotes LUAD progression by activating the JNK/c-JUN signaling pathway. Additionally, our ndings suggest that both GRP75 and HMGA1 are useful therapeutic targets for LUAD and that a high expression level of HMGA1 is an independent biomarker for predicting recurrence and a poor prognosis in stage I LUAD.