TCGA samples information and TLN1 expression analysis in pan-cancers and AML
The RNAseq data of the AML and the normal control samples are uniformly processed by the Toil process downloaded from UCSC XENA, and the RNAseq data in TPM (transcripts per million reads) format are analyzed and compared after log2 conversion. We found that TLN1 is abnormally highly expressed in 12 tumors including AML, compared with normal tissues (Figure1A and B). Then we used The Human Protein Atlas(https://www.proteinatlas.org/) to analyze the expression level of TLN1 in human blood cells, found that the RNA expression of TLN1 showed low cell-type specificity (Figure1C). Finally, we used UCSC XENA to analyze the RNA expression, exon expression and DNA methylation of TLN1 in the TCGA AML cohort(Figure1D). We found that DNA methylation level was significantly increased at the probe of cg14025149 and cg14095761(Figure1E).
The relationship between the expression of TLN1 and the prognosis and clinical characteristics of patients
We used survival package(3.2-10 version) for statistical analysis of survival data and survminer package [0.4.9 version] for visualization and drawing Kaplan-Meier survival curves. The high expression of TLN1 is related to poor overall survival(OS) in 7 tumors including AML(Figure2A-G). ROC analysis for AML patients shows that TLN1 was a significant prognostic marker for OS(Figure2H). Next, we downloaded the clinical characteristics of TCGA patients and divided them into high expression groups and low expression groups according to the median of TPM(Table1). We used univariate Cox regression analysis to evaluate prognostic variables related to OS. The univariate Cox model showed that high TLN1 expression is closely related to poor OS(HR=1.728 (95% Confidence interval:1.128−2.647)). In addition, age (>60 years), Cytogenetic risk(intermediate and poor) were also poor prognostic factors for AML patients(Figure2M). Therefore, we used the nomogram to show the correlation between TLN1 expression and prognosis. The Conformity Index (C-index) was 0.721(95% Confidence interval:0.693-0.748)(Figure2N). The results of the multiple hypothesis test (Dunn’s test) show that the groups of intermediate and poor are higher than the average of Favorable, the difference is statistically significant (P<0.05)( Figure2I). Mann-Whitney U test (Wilcoxon rank-sum test) results show that the expression level of TLN1 increased significantly when the age>60, the median of the difference between the two groups is 0.327(0.08-0.545), the difference is statistically significant (P=0.009)(Figure2K). Similarly, the expression of TLN1 was also related to the later FAB classifications and dead of OS events(Figure2J and L).
TLN1 related gene enrichment analysis
To investigate genes related to TLN1, we used LinkedOmics to analyze co-expressed genes of TLN1 in AML13. After screening with P<0.05 and |correlation coefficient|>0.5, we got 285 genes revealed a significant positive association with TLN1, while 165 genes indicated significantly negative associations(Figure3A-C). Then we used 450 genes related to TLN1 for GO|KEGG and GSEA enrichment analysis. GO enrichment analysis results showed that TLN1 may be involved in neutrophil mediated immunity, neutrophil activation and neutrophil degranulation etc(Figure3D and E). We also analyzed the bubble chart and chord chart for several main enrichment functions(Figure3F-H), the enrichment function notes are in supplementary 1. In addition, GSEA enrichment analysis results show that TLN1 may be related to multiple signaling pathways in hematological tumors, including tyrosine kinase receptor, PI3K/AKT/mTOR and FLT3 signaling pathway(Figure4A-K). Next , the analysis of cancer pathway activity showed that TLN1 may participate in the activation of EMT, PI3K/AKT, RAS/MAPK and RTK pathways(Figure4L). Complementary, drug sensitivity analysis showed TLN1 may be related to the resistance of PD-153035, austocystin-D and Afatinib. (5Z)-7-Oxozeaenol, AZD6482 and BX-795 may provide a therapeutic strategy for targeting TLN1(Figure4M and N).
PPI and immune infiltration analysis of TLN1
In the previous correlation analysis, MYH9 and FLNA showed a very significant correlation, and the correlation coefficients were 0.76(p<0.001) and 0.71(p<0.001). Therefore, we analyzed the expression and prognosis of MYH9 and FLNA in the TCGA AML cohort. The results showed that MYH9 was abnormally highly expressed in AML patients and both of MYH9 and FLNA were associated with poor prognosis(Figure5A). Using GeneMANIA to predict TLN1 and MYH9-related PPI network results showed that the functions were concentrated in blood coagulation, cell-cell junction organization, coagulation, hemostasis, homotypic cell-cell adhesion, platelet activation(Figure5B). We co-expression analysis of the top fifteen gens with the highest scores in the PPI network(Figure5C), and the expression of PIP5K1C ,ROCK1, S100A4, MY01A and WAC is abnormally increased in AML(Figure5D-H). This showed that TLN1 may work together with PIP5K1C ,ROCK1, S100A4, MY01A and WAC in AML. The results of immune infiltration analysis of TLN1 using R showed that TLN1 expression was significant associations with multiple immune cells including macrophages(p<0.001,r=0.428),Tem(p<0.001,r=0.426),iDC(p<0.001,r=0.421) etc(Figure5I). This means that TLN1 may be an important marker of immune infiltration in AML.
Expression of TLN1 in AML samples and cell lines
To further investigate the expression of TLN1 in AML, we verified the expression level of TLN1 in AML cell lines using qRT-PCR and Western blot(Figure6A and B). The result showed that the mRNA and protein expression levels of Talin1 were the highest in the acute monocytic leukemia cell line(THP-1). Next, we verified the protein expression level of Talin1 in the mononuclear cells of normal people and AML patients. The result showed that Talin1 expression was low or undetectable in normal people, and the expression level is abnormally increased in most AML patients(Figure6C and D). In addition, we used immunofluorescence staining to observe the localization of Talin1 in AML cell lines and the results showed that Talin1 is mainly located in the cell membrane and cytoplasm(Figure6E).
Silenced TLN1 can inhibit the proliferation of AML cells and promote differentiation
After transfecting THP-1 and HL60 cell lines with small interfering RNA(siRNA), the expression level of TLN1 was detected by qRT-PCR. As shown in Figure7A, both siRNA1 and siRNA2 could significantly reduce the expression of TLN1. Growth curve analysis showed that silenced TLN1 significantly reduced the proliferation ability of AML cells(Figure7E). Wright Giemsa compound staining results showed that silenced TLN1 leads to the transformation of AML cell nuclei from round to rod-shaped nuclei and paging nuclei and the nucleus of THP-1 appears pyknosis and deep staining(Figure7B). This means that silencing TLN1 could inhibit proliferation and promote the differentiation of AML cells. Flow cytometry was used to analyze the effect of TLN1 on the cell cycle, as shown in Figure 7C and D, the proportion of apoptotic cells of HL60 and THP-1 after TLN1 silence increased significantly. At the same time, the proportion of THP-1 cells in G0/G1 phase increased significantly, but this phenomenon was not found in HL60. We also tested the expression level of apoptosis protein silenced by TLN1, as shown in Figure7F, the expression of Caspase3 in TLN1 silenced group cells was significantly increased, but PARP did not change significantly. This means that the silencing of TLN1 may lead to apoptosis of AML cells. In summary, these results indicate that TLN1 participates in the progression of AML by regulating the proliferation, apoptosis, differentiation and cell cycle of AML cells.
TLN1 regulates the proliferation and differentiation of AML cells by Talin1/p-AKT/CREB signaling pathway
According to the enrichment analysis result of GSEA, TLN1 may be related to the PI3K/AKT signaling pathway. Therefore, we tested the expression levels of AKT and P-AKT after TLN1 silence. Our results showed that interfering with the expression of Talin1 reduced the expression level of P-AKT. Similarly, the expression level of CREB was also downregulated which was a signal molecule downstream of AKT and related to cell proliferation and differentiation(Figure7G). These results indicated that TLN1 may regulate the proliferation and differentiation of AML cells by regulating the expression of Talin1/p-AKT/CREB.