DLAT mRNA was abnormally expressed in pan-cancer
To observe DLAT mRNA expression levels in human cancers, we first obtained RNAseq data from TCGA database. From Fig.1A, it can be seen that DLAT expression levels varied in different types of cancers. Compared to normal tissues, DLAT expression was lower in 8 types of cancer, including breast invasive carcinoma (BRCA), head and neck squamous cell carcinoma (HNSC), colon adenocarcinoma (COAD), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), rectal cancer (READ), prostate cancer (PRAD), and thyroid cancer (THCA). While, DLAT expression was higher in 7 types of cancer, including cholangiocarcinoma (CHOL), kidney chromophobe (KICH), lung adenocarcinoma (LUAD), liver hepatocellular carcinoma (LIHC), lung squamous cell carcinoma (LUSC), uterine corpus endometrial carcinoma (UCEC), stomach cancer (STAD), compared to normal tissues. Considering that normal samples were fewer in TCGA database, we further verified DLAT expression in 33 cancers by intergrating TCGA and GTEx databases. As shown in Fig. 1B, it can be found that DLAT was abnormally expressed in 27 types of cancer. We found that DLAT was highly expressed in 22 types of cancer, including BRCA, CHOL, COAD, diffuse large B-cell Lymphoma (DLBCL), esophageal carcinoma (ESCA), glioblastoma (GBM), KICH, lower grade glioma (LGG), LIHC, LUAD, LUSC, ovarian cancer (OV), pancreatic cancer (PAAD), PRAD, READ, skin cutaneous melanoma (SKCM), STAD, testicular germ cell tumors (TGCT), THCA, thymoma (THYM), UCEC, uterine carcinosarcoma (UCS). However, it was lowly expressed only in 5 types of cancer, including adrenocortical carcinoma (ACC), bladder cancer (BLCA), HNSC, KIRC, and acute myeloid leukemia (LAML). Next, we analyzed DLAT expression in paired tumor samples, and found that DLAT was lowly expressed in COAD, KIRC, KICH, KIRP, LUSC, STAD, THCA, while highly expressed in CHOL, HNSC, LIHC, LUAD (Fig. 1C).
DLAT protein levels in pan-cancer
From the analysis of CPTAC samples on UALCAN platform, DLAT protein expression in BRCA, HNSC, GBM, KIRC and PAAD were significantly decreased, while the expression in UCEC were significantly increased (Fig. 2). In addition, we further analyzed the IHC staining patterns of DLAT in some tumors using the HPA database, and the results indicated that DLAT protein levels were consistent with the UALCAN results (Fig. 2).
DLAT was a prognostic biomarker in pan-cancer
Given that DLAT was aberrantly expressed in numerous cancer types, we speculated whether DLAT influences the survival of cancer patients. Hence, the correlation between DLAT expression and cancer patients’ OS, DSS and PFI were explored. Our results revealed that DLAT expression could significantly influence OS in patients with BRCA, COAD, esophageal adenocarcinoma (ESAD), HNSC, KICH, KIRC, KIRP, LGG, LIHC, PAAD, READ, STAD, and uveal melanoma (UVM) (Fig. 3A). According K-M survival curves, it can be found that low expression of DLAT predicted favorable OS in patients with BRCA, ESAD, HNSC, KICH, LGG, LIHC, PAAD, UVM. However, in patients with COAD, KIRC, KIRP, READ, STAD, those with low DLAT expression had unfavorable OS (Fig. 3B-N).
Moreover, the relationship between DSS and DLAT expression in human cancers were presented in Fig.4A. We demonstrated that ACC, BRCA, LGG, LIHC, PAAD, PRAD, sarcoma (SARC), and UVM patients with DLAT over-expression exhibited a shorter DSS. However, COAD, KIRC, KIRP, STAD patients with DLAT over-expression exhibited a longer DSS (Fig. 4B-M).
From Fig. 5A, it can be seen that DLAT expression had significantly effects on PFI in 13 cancer types, including ACC, BLCA, COAD, HNSC, KIRC, KIRP, LGG, LIHC, LUSC, PAAD, READ, STAD, and UVM. DLAT up-regulation exhibited a poor PFI in patients with ACC, BLCA, HNSC, LGG, LIHC, PAAD, UVM. However, in patients with COAD, KIRC, KIRP, LUSC, READ, STAD, those with DLAT up-regulation expression exhibited a good PFI (Fig. 5B-N).
DLAT was a diagnostic biomarker in pan-cancer
To determine the diagnostic value of DLAT, we performed ROC curves to calculate AUC value in cancers. According to the value of AUC, DLAT expression had high value for LAML, PAAD, CHOL, and KIRC early diagnosis (AUC: > 0.9), relative value for GBM, OV, LGG, LUAD, DLBCL, HNSC, LIHC, KIRP, THYM, READ, LUSC, STAD, KICH early diagnosis (AUC: 0.7-0.9),low value for ESAD, THCA, UCS, ACC, UCEC, ESCA, COAD, PRAD, TGCT, CESC, SKCM, BRCA, BLCA early diagnosis (AUC: 0.5-0.7) (Fig. 6).
DLAT genetic alteration in pan-cancer
We found a high DLAT amplification rate in UCS (1.75%), a high mutation rate in UCEC (2.65%), the highest incidence of “deep deletion” in TGCT (3.36%) (Fig. 7A). Missense and truncation were the main types of DLAT mutations (Fig. 7B). V242Wfs*5/E239G alteration in the 3D structure of DLAT was shown in Fig. 7C.
DLAT promoter methylation levels in pan-cancer
Increasing evidence have demonstrated that elevated/decreased gene methylation contribute to tumors development through inhibiting/activating the oncogenes expression. We demonstrated that aberrant promoter methylation of DLAT was involved in 13 cancer types. In BRCA, KIRC, LIHC, LUAD, PCPG, and SARC, DLAT promoter methylation levels were significantly decreased (Fig. 8A). While, elevated DLAT promoter methylation levels were found in BLCA, CESC, COAD, ESCA, HNSC, PRAD, and READ (Fig. 8B).
DLAT was associated with immune cell infiltration in pan-cancer
Extensive studies have illustrated that the infiltration of immune cells is related to the malignant biological behavior of tumors and can affect the prognosis of cancer patients. Therefore, the correlation between DLAT and 24 kinds of immune cell markers was investigated. From Fig. 9, the results revealed that DLAT was significantly associated with immune cell infiltration in the vast majority of tumors. Especially, we found that DLAT was positively associated with immune cell markers in GBM, including DC, B cells, macrophages, T cells, mast cells, NK cells, and so on.
Functional analysis of DLAT at single cell level
The results of CancerSEA database provided different functional states of DLAT in 17 tumors, including acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), Chronic myelogenous leukemia (CLL), GBM, glioma, astrocytoma (AST), high-grade glioma (HGG), oligodendroglioma (ODG), LUAD, non-small cell lung cancer (NSCLC), melanoma (MEL), renal cell carcinoma (RCC), BRCA, head and neck squamous cell carcinoma (HNSCC), colorectal cancer (CRC), retinoblastoma (RB), and uveal menaloma (UM) (Fig. 10A). Fig. 10B showed the positively relationship between DLAT expression and DNA damage and stemness in ALL. The positively relationship between DLAT expression and DNA damage and stemness in BRCA was displayed in Fig. 10C. In RB, DLAT expression was positively related to angiogenesis, differentiation, inflammation, negatively related to DNA repair, cell cycle, DNA damage (Fig. 10D). In UM, DLAT expression was negatively related to a variety of cancer-related functional states, including DNA repair, DNA damage, apoptosis, and metastasis (Fig. 10E). In addition, single-cell level LIPT1 expression profiles from ALL, BRCA, RB and UM were shown by T-SNE diagrams ((Fig. 10F-I).
Functional enrichment analysis of DLAT-related genes in pan-cancer
PPI network and functional enrichment analyses were performed to further explore role of DLAT in tumorigenesis. As shown in Fig. 11A, 10 proteins interacting with DALT were identified through STRING website. Next, we used GEPIA2.0 platform to obtain the top 100 DLAT-related genes in pan-cancer (Table. S1). The top 10 DLAT-related genes, including C11ORF57, CUL5, NDUFS1, PAFAH1B2, PDE12, RBM7, SDHD, TFAM, UBE4A, and ZW10, were significantly positively correlated with DLAT expression (Fig. 11B and C). The GO and KEGG analyses for the top 100 DLAT-related genes indicated that DLAT may play critical roles in mitochondrial matrix, coated vesicle and ribonucleoprotein granule (Fig. 11D).