Page 86 - GPD-2-1
P. 86

Gene & Protein in Disease                                                 A pan-cancer analysis of HMGB1




             A                                                    C














             B















             D












            Figure 1. Expression and protein level of HMGB1 in human tumors. (A) Expression level of HMGB1 in TCGA tumors versus adjacent tissues (if available)
            as visualized by TIMER2. *P < 0.05; **P < 0.01; ***P < 0.001. (B) Box plot representation of HMGB1 expression level comparison in CHOL, COAD, DLBC,
            GBM, LGG, PAAD, READ, STAD, and THYM (TCGA project) relative to the corresponding normal tissues (GTEx database). *P < 0.05. (C) Total protein
            level of HMGB1 in normal tissue and primary COAD, GBM, LIHC, OV, BRCA, UCEC, LUAD, and PAAD. Protein data was extracted and analyzed using
            CPTAC. **P < 0.01; ***P < 0.001. (D) Stage-dependent expression level of HMGB1. Major pathological stages (stage Ⅰ, stage Ⅱ, stageⅢ , and stage Ⅳ) of
            ACC, LIHC, SKCM, and THCA were assessed and compared using TCGA data. Expression levels are expressed in Log 2 (TPM + 1).
            HMGB1: High mobility group box 1; TCGA: The Cancer Genome Atlas; TIMER2: Tumor immune estimation resource, version 2; CHOL: Cholangiocarcinoma;
            COAD: Colon adenocarcinoma; DLBC: Diffuse large B-cell lymphoma; GBM: Glioblastoma multiforme; LGG: Lower grade glioma; PAAD: Pancreatic
            adenocarcinoma; READ: Rectum adenocarcinoma; STAD: Stomach adenocarcinoma; THYM: Thymoma; GTEx: Genotype-tissue expression; LIHC: Liver
            hepatocellular carcinoma; SKCM: Skin cutaneous melanoma; THCA: Thyroid carcinoma; BRCA: Breast cancer; UCEC: Uterine corpus endometrial
            carcinoma; LUAD: Lung adenocarcinoma; CPTAC: Clinical Proteomic Tumor Analysis Consortium.

            “amplification” type of CNA, with a frequency of about   3D  structure  of  HMGB1  protein  (Figure  3C).  Besides,
            ~4% (Figure  3A). It is worthwhile noting that all ACC   we searched the underlying association between certain
            cases with genetic variation (about ~1% frequency) had   genetic variations of  HMGB1 and the clinical survival
            HMGB1 copy number deletion (Figure  3A).  Figure  3B   prognosis  of  patients  with  different  cancer  types.  We
            shows the types, sites and number of cases of  HMGB1   studied and collected information of various tumor types,
            gene variation. We have not distinguished the main   and the data in  Figure  3D showed that UCEC patients
            types of genetic changes, and their locations appear to be   with  HMGB1  variation  had  better  prognosis  in  terms
            sporadic, with some belonging to the HMG-box domain.   of progression-free survival (PFS) (P < 0.05) than those
            For example, a truncating mutation, R163*/Q alteration,   without  HMGB1 variations. However, no significant
            in the HMG-box domain, was only detected in 2  cases   differences were observed in OS (P  = 0.0505), disease-
            of UCEC and 1 case of GBM. The R163*/Q site is in the   specific survival (DSS) (P > 0.05), and DFS (P > 0.05).


            Volume 2 Issue 1 (2023)                         4                         https://doi.org/10.36922/gpd.301
   81   82   83   84   85   86   87   88   89   90   91