Shao Wei, Gu Jinping, Huang Caihua, Liu Dan, Huang Huiying, Huang Zicheng, Lin Zhen, Yang Wensheng, Liu Kun, Lin Donghai, Ji Tianhai
Chenggong Hospital and College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China.
Mol Cancer. 2014 Aug 27;13:197. doi: 10.1186/1476-4598-13-197.
Ambiguity in malignant transformation of glioma has made prognostic diagnosis very challenging. Tumor malignant transformation is closely correlated with specific alterations of the metabolic profile. Exploration of the underlying metabolic alterations in glioma cells of different malignant degree is therefore vital to develop metabolic biomarkers for prognosis monitoring.
We conducted (1)H nuclear magnetic resonance (NMR)-based metabolic analysis on cell lines (CHG5, SHG44, U87, U118, U251) developed from gliomas of different malignant grades (WHO II and WHO IV). Several methods were applied to analyze the (1)H-NMR spectral data of polar extracts of cell lines and to identify characteristic metabolites, including principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), fuzzy c-means clustering (FCM) analysis and orthogonal projection to latent structure with discriminant analysis (OPLS-DA). The expression analyses of glial fibrillary acidic protein (GFAP) and matrix metal proteinases (MMP-9) were used to assess malignant behaviors of cell lines. GeneGo pathway analysis was used to associate characteristic metabolites with malignant behavior protein markers GFAP and MMP-9.
Stable and distinct metabolic profiles of the five cell lines were obtained. The metabolic profiles of the low malignancy grade group (CHG5, SHG44) were clearly distinguished from those of the high malignancy grade group (U87, U118, U251). Seventeen characteristic metabolites were identified that could distinguish the metabolic profiles of the two groups, nine of which were mapped to processes related to GFAP and MMP-9. Furthermore, the results from both quantitative comparison and metabolic correlation analysis indicated that the significantly altered metabolites were primarily involved in perturbation of metabolic pathways of tricarboxylic acid (TCA) cycle anaplerotic flux, amino acid metabolism, anti-oxidant mechanism and choline metabolism, which could be correlated with the changes in the glioma cells' malignant behaviors.
Our results reveal the metabolic heterogeneity of glioma cell lines with different degrees of malignancy. The obtained metabolic profiles and characteristic metabolites are closely associated with the malignant features of glioma cells, which may lay the basis for both determining the molecular mechanisms underlying glioma malignant transformation and exploiting non-invasive biomarkers for prognosis monitoring.
胶质瘤恶性转化的不确定性使得预后诊断极具挑战性。肿瘤恶性转化与代谢谱的特定改变密切相关。因此，探索不同恶性程度的胶质瘤细胞潜在的代谢改变对于开发用于预后监测的代谢生物标志物至关重要。
我们对源自不同恶性级别（世界卫生组织II级和IV级）胶质瘤的细胞系（CHG5、SHG44、U87、U118、U251）进行了基于氢核磁共振（NMR）的代谢分析。应用多种方法分析细胞系极性提取物的氢核磁共振光谱数据并鉴定特征性代谢物，包括主成分分析（PCA）、偏最小二乘判别分析（PLS-DA）、模糊c均值聚类（FCM）分析和带有判别分析的正交投影到潜在结构（OPLS-DA）。胶质纤维酸性蛋白（GFAP）和基质金属蛋白酶（MMP-9）的表达分析用于评估细胞系的恶性行为。基因通路分析用于将特征性代谢物与恶性行为蛋白标志物GFAP和MMP-9相关联。
获得了五个细胞系稳定且独特的代谢谱。低恶性级别组（CHG5、SHG44）的代谢谱与高恶性级别组（U87、U118、U251）的代谢谱明显不同。鉴定出17种可区分两组代谢谱的特征性代谢物，其中9种与GFAP和MMP-9相关的过程有关。此外，定量比较和代谢相关性分析的结果均表明，显著改变的代谢物主要参与三羧酸（TCA）循环回补通量、氨基酸代谢、抗氧化机制和胆碱代谢等代谢途径的扰动，这可能与胶质瘤细胞恶性行为的变化相关。
我们的结果揭示了不同恶性程度的胶质瘤细胞系的代谢异质性。所获得的代谢谱和特征性代谢物与胶质瘤细胞的恶性特征密切相关，这可能为确定胶质瘤恶性转化的分子机制以及开发用于预后监测的非侵入性生物标志物奠定基础。
