Expression and clinical significance of HMGA1 in pCCA
Transcriptome sequencing profiles in eight pCCA tissues and paired tumor-adjacent bile duct tissues were used to identify differentially expressed genes in pCCA (BioProject accession PRJNA517030 and PRJNA547373). According to the criteria set as fold change > 4 and Padj༜0.10, 180 genes were selected(Supplemental Table 5). Among the HMG family, only HMGA1 was significantly upregulated in pCCAs compared with adjacent bile duct tissues (Fig. 1A). Moreover, qRT-PCR with 18 pCCA pairs (Fig. 1B) and WB with four pCCA pairs further confirmed HMGA1 upregulation in pCCA(Fig. 1C). In a retrospective cohort of patients with pCCA radical resection (n = 106), the expression and localization of HMGA1 were detected with IHC. HMGA1 was mainly expressed in the nucleus (Fig. 1D), consistent with its function as a transcription factor.
Moreover, high expression of HMGA1 was significantly associated with positive lymphatic invasion and advanced TNM stage (Fig. 1E), indicating that HMGA1 may promote the invasion of pCCA. In overall survival curves, high HMGA1 expression was correlated with unfavorable prognosis (Fig. 1F; Supplementary Table 6), suggesting that HMGA1 may be a prognostic biomarker of pCCA. Cox-regression hazard models confirmed that HMGA1 tended to be an independent prognostic biomarker of pCCA, but the statistical significance is not that notable (P = 0.072) (Supplementary Table 6).
HMGA1 promoted the proliferation, invasion, stemness, and EMT in pCCA cells
Next, we measured HMGA1 expression in various biliary cell lines including normal biliary epithelium cell line HIBEpiC, the pCCA cell lines QBC-939 and FRH-0201, iCCA cell lines RBE and HCCC-9810, gallbladder carcinoma cell lines GBC-SD, NOZ and SGC-996. HIBEpiC cells showed significantly lower HMGA1 expression, whereas the expression of HMGA1 was increased in all these cell lines(Fig. 1G, Supplementary Fig. 1A). In both QBC-939 and FRH-0201 cells, HMGA1 was silenced with two independent shRNAs or overexpressed with a lentivirus carrying the HMGA1 cDNA (Supplementary Fig. 1B and 1C). CCK8 and colony formation assays demonstrated that HMGA1 knockdown significantly impaired proliferation, whereas HMGA1 overexpression promoted pCCA proliferation (Fig. 1H and 1I, Supplementary Fig. 2A). Stable QBC-939 cells with HMGA1 knockdown or overexpression were injected subcutaneously to establish xenografts in mice. The tumor volume and weight of xenografts were extensively decreased by HMGA1 knockdown and increased by HMGA1 overexpression (Fig. 1J and 1K, Supplementary Fig. 2B). In addition, wound healing and transwell assays demonstrated that HMGA1 promoted the migration and invasion of QBC-939 and FRH-0201 cells (Fig. 1L and 1M, Supplementary Fig. 2C and 2D). All above results indicated that HMGA1 had extensive influences on pCCA progression including proliferation, migration and invasion.
To explain this multiple functions of HMGA1 on pCCA progression, the effects of HMGA1 on cell stemness and EMT were investigated because previous study suggested HMGA1 is an important factor involved in cell stemness and EMT[13]. Sphere-formation assays showed that HMGA1 upregulation increased pCCA stemness, whereas HMGA1 downregulation suppressed stemness (Fig. 1N, Supplementary Fig. 2E). Moreover, E-cadherin expression was decreased, and other EMT biomarkers including N-cadherin, Vimentin, Snail, Twist-1, and Claudin-1, were upregulated following HMGA1 overexpression, and downregulated following HMGA1 knockdown(Fig. 1O and 1P). These results indicated that HMGA1 played important roles in stemness and EMT of pCCA, which thus influenced proliferation, migration and invasion.
HMGA1 promoted the transcription and expression of TRIP13
In mRNA sequencing of eight pairs of pCCA and normal bile duct tissues, 180 genes were up-regulated (Fig. 2A, Supplementary Table 5). In previous study, a total of 21 proteomic signatures regulated by HMGA1 in breast cancer were reported, and three of them (KIFC1, LRRC59, and TRIP13) were verified to promote breast cancer progression[31]. Interestingly, TRIP13, was identified by both our mRNA sequencing and previous proteomic HMGA1-linked signatures (Fig. 2A, Supplementary Table 5). The mRNA levels of KIFC1, LRRC59, and TRIP13 were evaluated using 36 cases of CCA from The Cancer Genome Atlas (TCGA) database(https://tcga-data.nci.nih.gov/tcga/), and their correlations with HMGA1 were analyzed (Fig. 2B). KIFC1 and TRIP13 showed positive correlation with HMGA1 of the 36 CCAs, but qRT-PCR showed that only TRIP13 expression was regulated by HMGA1 in QBC-939 (Fig. 2C). Our qRT-PCR results with 18 pCCA tissues also supported the strong positive correlation between HMGA1 and TRIP13 (Fig. 2D). TRIP13 expression was detected by IHC in 106 cases in pCCA TMA (Fig. 2E). The IHC score of TRIP13 was significantly associated with the IHC score of HMGA1(Fig. 2F), and patients with high HMGA1 expression had high TRIP13 expression (Fig. 2G). In QBC-939 cells, regulation of HMGA1 expression led to corresponding changes of TRIP13 (Fig. 2H). Finally, luciferase assays demonstrated that HMGA1 promoted the transcription of TRIP13 in pCCA cells(Fig. 2I) and 293T cells(Fig. 2J). All above results suggested that HMGA1 induced TIRP13 expression via promoting its transcription.
TRIP13 promoted cancer progression and was correlated with poor prognosis in pCCA
TRIP13 expression was highest in QBC-939 cells among the detected biliary cell lines including HIBEpiC, QBC-939, FRH-0201, RBE, HCCC-9810, GBC-SD, NOZ and SGC-996 (Fig. 3A). Evaluation of TRIP13 expression by qRT-PCR in 18 pairs of pCCA tissues and WB in four pairs of pCCA tissues demonstrated that TRIP13 was upregulated (Fig. 3B and 3C). In the validation cohort, patients with high expression of TRIP13 had poorer prognoses than those with low expression(P = 0.019) (Fig. 3D). Intriguingly, high expression of both TRIP13 and HMGA1 was a more sensitive prognostic factor than TRIP13 or HMGA1 alone (P = 0.0002) (Fig. 3D). Multivariate analysis also identified TRIP13 as an independent prognostic biomarker of pCCA (hazard ratio = 1.95, P = 0.046; Supplementary Table 6). Importantly, TRIP13 was significantly associated with TNM stage and tended to be associated with lymphatic invasion, similar to HMGA1 (Supplementary Table 7). CCK8 assays demonstrated that TRIP13 promoted the proliferation of pCCA cells (Fig. 3E). Wound healing and transwell assays suggested that TRIP13 was required in pCCA cell migration and invasion (Fig. 3F and 3G, Supplementary 3A and 3B). Similar to HMGA1, TRIP13 was essential for stemness and the EMT of pCCA cells (Fig. 3H-3J, Supplementary 3C). Taken together, these results suggested that TRIP13 promoted the progression of pCCA.
TRIP13 was required in HMGA1-induced pCCA progression
To detect the role of TRIP13 in HMGA1-induced pCCA progression, TRIP13 was silenced in HMGA1-overexpressing cells. CCK8 and colony formation assays demonstrated that TRIP13 knockdown attenuated HMGA1-induced proliferation in QBC-939 and FRH-0201 cells (Fig. 4A and 4B, Supplementary Fig. 4A). Xenografts were established with HMGA1-overexpressing stable cells with or without TRIP13 knockdown, showing that TRIP13 knockdown significantly reduced tumor volume and weight which were increased by HMGA1 overexpression (Fig. 4C and 4D). The migration and invasion of HMGA1-overexpressing QBC-939 and FRH-0201 cells were impaired after TRIP13 knockdown (Fig. 4E and 4F, Supplementary Fig. 4B). Stable QBC-939 cells with HMGA1 overexpression and/or TRIP13 knockdown were injected into the tail vein, and metastases to the liver was detected with HE staining (Fig. 4G). HMGA1 overexpression increased the number of metastatic lesions, whereas silencing of TRIP13 neutralized this effect (Fig. 4H). 3D sphere formation and EMT biomarker expression showed that TRIP13 knockdown significantly impaired HMGA1-induced cell stemness and the EMT (Fig. 4I and 4J, Supplementary Fig. 4C). All these results indicated that HMGA1 promoted the stemness and EMT by elevating TRIP13 expression in pCCA cells.
FBXW7 suppressed TRIP13-induced progression by degrading c-Myc
We previously reported that F-box/WD repeat-containing protein 7(FBXW7) suppressed the stemness and EMT of CCA[32], and a recent study proposed that FBXW7 expression was inhibited by TRIP13 in glioblastoma[33], so we further investigated the correlation between FBXW7 and TRIP13 in pCCA progression. WB, qPCR and luciferase assay showed that TRIP13 knockdown significantly increased the transcription and expression of FBXW7 in both QBC-939 and FRH-0201(Fig. 5A and 5B, Supplementary Fig. 5A and 5B). TRIP13 and FBXW7 mRNA levels in the HMGA1-silenced and HMGA1-overexpressed xenografts(Fig. 1J and Fig. 4C) were detected with qRT-PCR, reflecting that TRIP13 and FBXW7 are downstream effectors of HMGA1(Fig. 6). Moreover, the FBXW7 knockdown facilitated the proliferation, migration and invasion of pCCA cells, which was attenuated by TRIP13 knockdown(Fig. 5C-5E, Supplementary Fig. 5C). Biomarkers of stemness or EMT, and sphere formation assay were performed after silencing TRIP13 or FBXW7 using WB and qRT-PCR(Fig. 5F-5H Supplementary Fig. 5D and 5E). Consequently, TRIP13 knockdown attenuated stemness and EMT of pCCA cells, while FBXW7 knockdown reversed this tendency, indicating that FBXW7 was involved in the TRIP13-induced stemness and EMT.
c-Myc is a well-known onco-protein involved in stemness, and also a target of FBXW7 for ubiquitination. In QBC-939 and FRH-0201, c-Myc expression was elevated by FBXW7 knockdown and reduced by TRIP13 knockdown(Fig. 6A, Supplementary Fig. 7A), but qRT-PCR implicated that c-Myc mRNA was not influenced(Fig. 6B), suggesting that c-Myc degradation, instead of transcription, was affected by TRIP13 and FBXW7. Incubation in the ubiquitination inhibitor MG132(10uM) extensively eliminated the FBXW7-induced degradation of c-Myc (Fig. 6C, Supplementary Fig. 7B). Additionally, FBXW7 knockdown substantially decreased ubiquitinated c-Myc in HA-Ubiquitin-overexpressing QBC-939(Fig. 6D). Collectively, TRIP13 promoted pCCA stemness and EMT by suppressing FBXW7 transcription and thus stabilizing c-Myc.
HMGA1-TRIP13 axis promotes stemness and EMT in a positive feedback pathway dependent on c-Myc
With the transcription-factor-predicting software(Jaspar software), we found that c-Myc was predicted to promote the transcription of HMGA1 and TRIP13 so we further investigated the role of c-Myc in HMGA1 and TRIP13 expression. WB(Fig. 6E and 6F) and qRT-PCR (Supplementary Fig. 8A-8C) showed that both 10058-F4(a c-Myc inhibitor) and c-Myc knockdown significantly decreased expression of HMGA1 and TRIP13. Luciferase assay validated that c-Myc overexpression increased the transcription of HMGA1 and TRIP13, and c-Myc knockdown had contrary effects(Fig. 6G). These results implicated that c-Myc was able to induce the transcription of HMGA1 and TRIP13. Moreover, we knocked down TRIP13 in QBC-939, and demonstrated that TRIP13 can also regulate HMGA1 expression(Fig. 6H). However, it was interesting to note that TRIP13 expression was decreased almost at the same time of HMGA1 knockdown, while HMGA1 expression was attenuated about 12 hours later than TRIP13 knockdown(Fig. 6H). This may be explained by that TRIP13 regulated HMGA1 expression by stabilizing c-Myc, requiring more time than that HMGA1 directly regulated TRIP13 transcription. Combined with previous results that HMGA1-TRIP13 axis stabilized c-Myc, we postulated that HMGA1 had a positive feedback loop to amplify its biological effect depending on c-Myc.
Wnt-β-catenin pathway is a well-accepted activator of c-Myc, so Wnt3a(100ng/ml) was used to incubate QBC-939 for 12 hours to stimulate Wnt signaling. The recently developed small-molecule inhibitor of TRIP13, DCZ0415(10uM)[34], HMGA1 inhibitor Netropsin(10ug/ml), and 10058-F4(10uM) were used to incubate QBC-939 cells. Interestingly, any inhibitor can decrease the expression of c-Myc, HMGA1 and TRIP13(Fig. 7A, Supplementary Fig. 7D), suggesting that these three factors were in the same positive feedback loop. Moreover, the inhibitors of HMGA1, TRIP13 and c-Myc inhibited the stemness and EMT of QBC-939 cells with Wnt3a stimulation (Fig. 7B-7D, Supplementary Fig. 7E), and suppressed pCCA migration and invasion(Fig. 7E and 7F, Supplementary Fig. 7F). All these results suggested that c-Myc promotes transcription and expression of HMGA1 and TRIP13, and implicated that HMGA1-TRIP13 axis facilitates pCCA progression in a c-Myc-dependent positive feedback loop.
In our previous study, we showed that TCF family, important component of Wnt-β-catenin signaling, can induce c-Myc expression and promote pCCA progression[9], therefore we detected the correlation between HMGA1 and TCF family, which showed that HMGA1 regulated TCF4/TCF7/LEF1 expression and their downstream effector c-Myc(Supplementary Fig. 9). This result suggests that HMGA1 has multiple crosslinks with Wnt-β-catenin-Myc signaling, which may be another positive feedback loop to amplify cell stemness and EMT in pCCA.