DNAAF5 promotes hepatocellular carcinoma malignant progression by recruiting USP39 to improve PFKL protein stability

Abstract

Purposes: Dynein axonemal assembly factor 5 (DNAAF5) is the transcription factor of regulating the cytoskeleton and hydrodynamic protein complex assembly, however, it was not well elucidated in the malignant progression of hepatocellular carcinoma (HCC).

Methods: We investigated the role of DNAAF5 in hepatocellular carcinoma by using multiple groups of clinical tissues combined with data from the TCGA database. Then we overexpressed DNAAF5 in hepatocellular carcinoma tumor tissues, which correlates with poor patient survival outcomes. Furthermore, we constructed stable cell lines of HCC cells to confirm the cancer-promoting effects of DNAAF5 in hepatocellular carcinoma. To explore the mechanisms of DNAAF5, transcriptome sequencing combined with mass spectrometry was also performed, which showed that DNAAF5 affects its downstream signaling pathway by interacting with PFKL and that DNAAF5 regulates PFKL protein stability by recruiting the deubiquitination protein, USP39. To corroborate these findings, the same series of tissue microarrays were used to confirm correlations between DNAAF5 and PFKL expressions. In animal experiments, DNAAF5 also promoted the proliferation of HCC cells.

Results: We found that DNAAF5 expressions were markedly higher in HCC tissues, compared to the adjacent normal tissues. Increased levels of DNAAF5 were associated with significantly worse prognostic outcomes for HCC patients. Cell function experiments showed that HCC cells of overexpressing DNAAF5 exhibited faster proliferation rates, stronger clone formation abilities and higher drug resistance rates. However, tumor cell proliferation rates and colony formation were significantly decreased after DNAAF5 knockout, accompanied by an increase in sensitivity to sorafenib. In addition, the results of our study showed that DNAAF5 accelerates PFKL protein deubiquitination by recruiting USP39 in HCC cells. Furthermore, The overexpression of DNAAF5 could promote HCC cell proliferation in vivo and in vitro, whereas USP39 knockdown inhibited this effect. Overall, DNAAF5 serves as a scaffold protein to recruit USP39 to form a ternary complex by directly binding the PFKL protein, thereby improving the stability of the latter, which promotes the malignant process of hepatocellular carcinoma.

Conclusions: These findings revealed DNAAF5 was negatively correlated with the prognosis of patients with hepatocellular carcinoma. It underlying mechanism showed that DNAAF5 directly binds PFKL and recruits the deubiquitinated protein (USP39) to improve the stability of the PFKL protein, thus enhancing abnormal glycolysis in HCC cells.

Keywords: DNAAF5; HCC; PFKL protein; glycolysis; prognosis.

Figure 1

DNAAF5 is highly expressed in hepatocellular carcinoma and negatively correlates with prognosis. (A) Expression of DNAAF5 in 12 paired human HCC tumor tissues (T) and matched paracancerous tissues (N) as determined by western blotting assays. (B) The mRNA expression of DNAAF5 in hepatocellular carcinoma tissues and in non-tumor tissues adjacent to the cancer as determined using qPCR assays. (C) IHC staining results showing the expression of DNAAF5 in cancer tissue and adjacent non-cancerous tissues. Representative staining images of DNAAF5 are shown (left), and high DNAAF5 expression was negatively correlated with patient prognosis (right). (D) Analysis of sequencing data of hepatocellular carcinoma in TCGA database indicating higher expression of DNAAF5 in tumor tissues compared to adjacent non-cancerous tissues (left). The expression of DNAAF5 in hepatocellular carcinoma was negatively correlated with patient prognosis (middle). ROC curve (AUG>0.9) illustrating the accuracy of DNAAF5 in predicting patient prognosis (right). **p < 0.01, ***p < 0.001 were the thresholds for statistical significance.

Figure 2

DNAAF5 promotes multiple malignant phenotypes in hepatocellular carcinoma cells. (A) Expression level of DNAAF5 in different HCC cell lines as determined by western blotting assays (left), qPCR results showing mRNA level of DNAAF5 in different HCC cell lines (right). (B) The expression of DNAAF5 reflecting the successful construction of stable cell lines as determined by western blotting assays. (C–E) Results of plate cloning assays (C), CCK8 assays (D), and sorafenib killing experiments (E), showed that DNAAF5 overexpression in HCC cells increased the degree of malignancy and vice versa. *p < 0.05, **p < 0.01, ***p < 0.001 were the thresholds for statistical significance; ns represents no statistical significance.

Figure 3

DNAAF5 regulate multiple signaling pathways in hepatocellular carcinoma. (A) Transcriptome sequencing results showed that DNAAF5 might regulate multiple signaling pathways (left) and affect the expression of a large number of genes (right). (B) The expression level of key enzymes involved in glycolysis as determined by western blotting assays. The expression PFKL was positively associated with expression level of DNAFF5, but the expression level of DNAAF5 did not affect the expression of HK2 and PKM2. (C) Intracellular expression level of three key transcription factors (p53, myc and HIF-1a) involved glycolysis was not altered significantly as revealed by western blotting assays. (D) Cytological results showing that overexpression of DNAAF5 increased glucose consumption and lactate production, and vice versa. ***p < 0.001.

Figure 4

DNAAF5 improves the stability of PFKL protein. (A) The interaction between DNAAF5 and PFKL proteins in Hep3B cells as determined using immunoprecipitation (IP) followed by immunoblotting. IgG was used as a negative control for immunoprecipitation. (B) Protein purification and pulldown results demonstrating direct interaction between DNAAF5 and PFKL. Results of GST- and His-pulldown experiments indicating that only GST-DNAAF5 directly interacted with His-PFKL, but not GST protein. (C) DNAAF5 overexpression significantly improved protein stability of PFKL in Hep3B cells, but knockout DNAAF5 enhanced the degradation of PFKL in Huh-7 cells (cells were exposed to cycloheximide (CHX 50 mg/ml) for 0, 3 or 6 h). (D) DNAAF5 overexpression significantly attenuated the ubiquitination of PFKL, while DNAAF5 knockdown significantly enhanced the ubiquitination of PFKL in HCC cells (cells were treated with MG132 (10 mM) for 6 h).

Figure 5

DNAAF5 accelerates the deubiquitination of PFKL protein by recruiting USP39. (A) The interaction between DNAAF5-PFKL and PFKL-USP39 proteins in Hep3B cells as determined using immunoprecipitation (IP) followed by immunoblotting. IgG served as the negative control for immunoprecipitation. (B) Protein purification and pulldown assays confirming the direct interaction between DNAAF5 and USP39. The GST- and His-pulldown experiments demonstrated that only GST-DNAAF5 directly interacted with His-USP39, but not GST protein. (C) Protein purification and pulldown assays results demonstrating the direct interaction between PFKL and USP39. The GST- and His-pulldown experiments showed that only GST-PFKL directly interacted with His-USP39, but not GST protein. (D) Upregulation of PFKL expression caused by changes in DNAAF5 expression was rescued by knock down of USP39 in Huh-7 and PLC/PRF-5 cells. (E) The CCK8 assays showed that DNAAF5 overexpression promoted the proliferation of tumor cells, whereas USP39 knockdown inhibited this effect in HCC cells. (F) The plate cloning assays showed that DNAAF5 overexpression promoted the growth of tumor cells, whereas USP39 knockdown inhibited this effect in Hep3B cells. ***p < 0.001.

Figure 6

DNAAF5 promotes the growth of hepatocellular carcinoma in vivo (A) IHC staining indicating a higher expression of PFKL in cancer samples compare to adjacent non-cancerous tissues, the chip shown is from the same group as in Figure 1 , and the same location was selected. PFKL expression gradually increased with increase in DNAAF5 expression, indicating strong correlation between DNAAF5 and Warburg effect. (B) The statistical results of immunohistochemical staining showing a significant positive correlation between DNAAF5 expression and PFKL expression level. (C) DNAAF5 overexpression promoted the growth of tumor xenografts in nude mice, whereas USP39 knockdown inhibited this effect. The left panel shows photographs of the tumors and the right panel shows the weight of tumors (scale bar: 2cm). (D) The immunohistochemical staining assays showed that DNAAF5 overexpression promoted the tumor proliferation in vivo assays, whereas USP39 knockdown inhibited this effect (scale bar: 10μm). ***p < 0.001; ns represents no statistical significance.