DEG Analysis
GSE 191215 data set includes three control samples and four different shRNA sequences targeting EZH2 ( three samples per sequence, a total of 12 samples) samples. Compare C1, G2, H2, H8 with the control group, and obtain the corresponding volcanic map (Fig. 2A-D) through the default GEO2R analysis and visualization, and use the Wayne diagram to intersect the up-regulated and down-regulated DEG of the four groups (Fig. 2E-F). We found that a total of 42 DEG had p values < 0.05 (31 of them were up and 11 were down).
Functional and Pathway Enrichment and Lasso regression Analysis
GO analysis showed that these DEGs were enriched in the following processes ( only the top 5 when the results were more than 5), in which the down-regulated genes: In BP nuclear-transcribed mRNA catabolic process, signal transduction in response to DNA damage, mRNA catabolic process, neuron apoptotic process, regulation of response to DNA damage stimulus. In CC dynein axonemal particle, RISC complex, RNAi effector complex, chaperone complex, integral component of mitochondrial membrane. In MF 5'-3' exonuclease activity, death receptor binding, telomerase RNA binding, mannosyltransferase activity, R-SMAD binding (Fig. 3A, Table 1).
Table 1
GO term enrichment analysis of down-regulated DEGs.
ONTOLOGY | ID | Description | GeneRatio | BgRatio | pvalue | p.adjust | qvalue | geneID | Count |
BP | GO:0000956 | nuclear-transcribed mRNA catabolic process | 2/7 | 121/18800 | 0.00084473 | 0.05671584 | 0.02633195 | DCP2/DDX5 | 2 |
BP | GO:0042770 | signal transduction in response to DNA damage | 2/7 | 174/18800 | 0.00173482 | 0.05671584 | 0.02633195 | BID/DDX5 | 2 |
BP | GO:0006402 | mRNA catabolic process | 2/7 | 233/18800 | 0.00308279 | 0.05671584 | 0.02633195 | DCP2/DDX5 | 2 |
BP | GO:0051402 | neuron apoptotic process | 2/7 | 241/18800 | 0.00329389 | 0.05671584 | 0.02633195 | BID/TMTC4 | 2 |
BP | GO:2001020 | regulation of response to DNA damage stimulus | 2/7 | 246/18800 | 0.00342922 | 0.05671584 | 0.02633195 | BID/DDX5 | 2 |
CC | GO:0120293 | dynein axonemal particle | 1/8 | 20/19594 | 0.0081381 | 0.058651 | 0.03268477 | DNAAF2 | 1 |
CC | GO:0016442 | RISC complex | 1/8 | 27/19594 | 0.01097271 | 0.058651 | 0.03268477 | DCP2 | 1 |
CC | GO:0031332 | RNAi effector complex | 1/8 | 27/19594 | 0.01097271 | 0.058651 | 0.03268477 | DCP2 | 1 |
CC | GO:0101031 | chaperone complex | 1/8 | 34/19594 | 0.01380023 | 0.058651 | 0.03268477 | DNAAF2 | 1 |
CC | GO:0032592 | integral component of mitochondrial membrane | 1/8 | 88/19594 | 0.03537585 | 0.07937415 | 0.04423327 | BID | 1 |
MF | GO:0008409 | 5'-3' exonuclease activity | 1/9 | 18/18410 | 0.00876713 | 0.06817168 | 0.03044349 | DCP2 | 1 |
MF | GO:0005123 | death receptor binding | 1/9 | 21/18410 | 0.01022165 | 0.06817168 | 0.03044349 | BID | 1 |
MF | GO:0070034 | telomerase RNA binding | 1/9 | 22/18410 | 0.01070607 | 0.06817168 | 0.03044349 | DCP2 | 1 |
MF | GO:0000030 | mannosyltransferase activity | 1/9 | 24/18410 | 0.01167428 | 0.06817168 | 0.03044349 | TMTC4 | 1 |
MF | GO:0070412 | R-SMAD binding | 1/9 | 24/18410 | 0.01167428 | 0.06817168 | 0.03044349 | DDX5 | 1 |
KEGG analysis showed that these DEGs were enriched in the following processes (only the top 5 when the results were more than 5), in which the down-regulated genes: Apoptosis - multiple species, Viral myocarditis, Platinum drug resistance, p53 signaling pathway, RNA degradation。(Fig. 3B, Table 2).
Table 2
KEGG pathway enrichment analysis of down-regulated DEGs.
ONTOLOGY | ID | Description | GeneRatio | BgRatio | pvalue | p.adjust | qvalue | geneID | Count |
KEGG | hsa04215 | Apoptosis - multiple species | 1/3 | 32/8164 | 0.01171434 | 0.09442006 | 0.06994079 | BID | 1 |
KEGG | hsa05416 | Viral myocarditis | 1/3 | 60/8164 | 0.02188904 | 0.09442006 | 0.06994079 | BID | 1 |
KEGG | hsa01524 | Platinum drug resistance | 1/3 | 73/8164 | 0.02658917 | 0.09442006 | 0.06994079 | BID | 1 |
KEGG | hsa04115 | p53 signaling pathway | 1/3 | 73/8164 | 0.02658917 | 0.09442006 | 0.06994079 | BID | 1 |
KEGG | hsa03018 | RNA degradation | 1/3 | 79/8164 | 0.02875337 | 0.09442006 | 0.06994079 | DCP2 | 1 |
Lasso regression analysis showed that there is a prognostic model of the prognostic value of these DEGs (Fig. 3C) and a Lasso variable trajectory plot based on the Lasso regression results (Fig. 3D).
Among the up-regulated genes, no DEGs were found to be enriched.
Clinical features
A total of 263 patients with sarcomas were included in this study. 144 female patients (54.8%) and 119 male patients (45.2%). 130 patients (49.5%) were under 60 years old. 199 patients (83.2%) had no tumor multifocal, while the remaining 40 patients (16.7%) had tumor multifocal. 21 patients (10.1%) had superficial tumor depth, while the other 188 patients (90%) had deep tumor depth. See Table 3 for more clinicopathological data of patients.
Table 3
Clinicopathological data of patients.
Characteristics | Low expression of TMTC4 | High expression of TMTC4 | P value |
n | 131 | 132 | |
Gender, n (%) | | | 0.2416 |
Female | 67 (25.5%) | 77 (29.3%) | |
Male | 64 (24.3%) | 55 (20.9%) | |
Age, n (%) | | | 0.0558 |
<= 60 | 57 (21.7%) | 73 (27.8%) | |
> 60 | 74 (28.1%) | 59 (22.4%) | |
Race, n (%) | | | 0.7222 |
Asian | 4 (1.6%) | 2 (0.8%) | |
White | 115 (45.3%) | 115 (45.3%) | |
Black or African American | 9 (3.5%) | 9 (3.5%) | |
Residual tumor, n (%) | | | 0.5660 |
R0 | 84 (35.7%) | 73 (31.1%) | |
R1 | 32 (13.6%) | 37 (15.7%) | |
R2 | 4 (1.7%) | 5 (2.1%) | |
Tumor multifocal, n (%) | | | 0.2598 |
No | 104 (43.5%) | 95 (39.7%) | |
Yes | 17 (7.1%) | 23 (9.6%) | |
Tumor necrosis, n (%) | | | 0.3684 |
No necrosis | 40 (21.9%) | 31 (16.9%) | |
Focal necrosis | 20 (10.9%) | 18 (9.8%) | |
Moderate necrosis | 28 (15.3%) | 34 (18.6%) | |
Extensive necrosis | 4 (2.2%) | 8 (4.4%) | |
Tumor depth, n (%) | | | 0.8001 |
Superficial | 11 (5.3%) | 10 (4.8%) | |
Deep | 93 (44.5%) | 95 (45.5%) | |
Metastasis, n (%) | | | 0.6711 |
No | 61 (34.1%) | 59 (33%) | |
Yes | 28 (15.6%) | 31 (17.3%) | |
Margin status, n (%) | | | 0.6525 |
Negative | 74 (34.7%) | 65 (30.5%) | |
Positive | 37 (17.4%) | 37 (17.4%) | |
Radiation therapy, n (%) | | | < 0.05 |
No | 81 (31.5%) | 98 (38.1%) | |
Yes | 47 (18.3%) | 31 (12.1%) | |
Expression level of TMTC4 and prognosis of sarcoma
Through reasonable grouping and analysis of GSE191215 data set, the intersection of the obtained results is taken. We can confirm that the expression level of TMTC4 in KRIB cells was significantly reduced after targeted knockout of IL11RA compared with the control group. Therefore, we have reason to believe that the low expression level of TMTC4 is closely related to the improvement of the prognosis of sarcoma.
Kaplan – Meier curve survival analysis showed that patients with low expression of TMTC4 showed better overall survival (OS) and disease specific survival (DSS) than patients with high expression of TMTC4(p = 0.025, p = 0.016, Fig. 4A-B).
Cox survival analysis showed that in univariate analysis, there was significant survival difference between TMTC4 high and low expression groups (risk ratio [HR], 1.579; 95% confidence interval CI [1.058–2.359]; p = 0.025) (Fig. 4C). In addition to the expression level of TMTC4, Metastasis, Residual tumor, Tumor multifocal and Margin status showed independent prognostic value in sarcomas.
Not only that, we further use Kaplan – Meier curve for subgroup analysis. he results showed that among female patients, patients with lower TMTC4 expression level had higher OS (p = 0.046), but excluding male patients (p = 0.143) (Fig. 5A-B). Among the patients with deeper tumor depth, the patients with lower TMTC4 expression level had higher OS (p = 0.042), but excluding the patients with shallower tumor depth (p = 0.469) (Fig. 5C-D). Among the patients with tumor metastasis, the patients with lower TMTC4 expression level had higher OS (p = 0.019), but excluding the patients without tumor metastasis (p = 0.592) (Fig. 5E-F). Among patients with negative tumor margin, patients with lower TMTC4 expression level had higher OS (p = 0.025), but excluding patients with positive tumor margin (p = 0.947) (Fig. 5G-H).
Construction predictive nomogram
In this study, Nomogram was constructed by using the expression levels of patients' Gender, Race, Age, Tumor multifocal, Tumor necrosis, Tumor depth and TMTC4 to predict the OS probability of sarcoma patients in 1, 2 and 3 years, which can provide useful information for individualized clinical evaluation and treatment strategies (Fig. 6A). The calibration chart shows that the 1-year, 2-year and 3-year OS probabilities predicted by our nomogram are basically consistent with the actual results (Fig. 6B-D). In conclusion, these results show that our nomogram can accurately predict the OS of patients with sarcoma.
GSEA
In order to understand the potential molecular mechanism of the prognostic value of TMTC4 in sarcoma, we conducted GSEA using transcriptome data from TCGA. The results of GSEA showed that in patients with low TMTC4 expression, IMMUNOREGULATORY and ANTIGEN_ ACTIVATES are enriched (Fig. 7A-B).
Evaluation of immune infiltration
The correlation between TMTC4 and 24 kinds of immune cells in patients with sarcoma was analyzed by lollipop diagram. The results showed that there was the strongest negative correlation with DC cells (P < 0.001) (Fig. 8A). Next, further analysis was conducted for DC cells to detect the difference of DC enrichment fraction between the high TMTC4 expression group and the low TMTC4 expression group (Fig. 8B-C). Use TIMER website to draw relevant KM curve (Fig. 8D). The results showed that when TMTC4 expression was low, the survival rate of high DC cell group was high (P < 0.001). When TMTC4 was highly expressed, there was no significant difference in the survival rate between the high DC cell group and the low DC cell group (P = 0.145).
Evaluation of methylation level of TMTC4 in sarcoma
According to the analysis of UALCAN website, the methylation level of TMTC4 promoter in tumor tissue is significantly higher than that in normal tissues adjacent to cancer (p < 0.001; Fig. 9A). In addition, further use of SMART website analysis can lead to Chromosomal distribution of the metal probes associated with TMTC4 (Fig. 9B) and Detailed genetic information of TMTC4 (Fig. 9C). Finally, MethSurv analysis showed that there were significant survival differences between patients with high and low TMTC4 methylation levels, and patients with high TMTC4 methylation levels had higher OS (Fig. 10A). In addition, Splitting of methylation by maxstat (Fig. 10B) and methylation profiles based on age (Fig. 10C) and sex (Fig. 10D) are obtained.
The expression level of TMTC4 in osteosarcoma cells
In order to study the expression level of TMTC4 in Osteosarcoma cells, we used osteoblast hFOB1.19 as the control group to detect the mRNA level of TMTC4 in Osteosarcoma cells MG-63, U2OS and Saos-2, respectively. The results showed that compared with the control group (hFOB1.19), the expression level of TMTC4 in the Osteosarcoma group (MG-63, U2OS, Saos-2) was significantly higher (p < 0.01) (Fig. 11), and the results we analyzed from the database were experimentally verified.
Inhibition of TMTC4 expression impedes cell proliferation
In order to verify the function of TMTC4 in Osteosarcoma cells again, we transfected TMTC4 overexpression plasmids, sh-TMTC4 and corresponding NCs into osteosarcoma cells. After confirming the successful transfection (Fig. 12A), we used CCK-8 experiment to detect its proliferation level. The results showed that knockout of TMTC4 would inhibit the level of cell proliferation, while overexpression of TMTC4 resulted in opposite effect (Fig. 12B).