Ustjanzew Arsenij, Nedwed Annekathrin Silvia, Sandhoff Roger, Faber Jörg, Marini Federico, Paret Claudia
Institute of Medical Biostatistics, Epidemiology and Informatics (IMBEI), University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, 55131, Germany.
Lipid Pathobiochemistry, German Cancer Research Center, Heidelberg, 69120, Germany.
Cancer Metab. 2024 Oct 24;12(1):29. doi: 10.1186/s40170-024-00358-y.
Glycosphingolipids (GSLs) are membrane lipids composed of a ceramide backbone linked to a glycan moiety. Ganglioside biosynthesis is a part of the GSL metabolism, which involves sequential reactions catalyzed by specific enzymes that in part have a poor substrate specificity. GSLs are deregulated in cancer, thus playing a role as potential biomarkers for personalized therapy or subtype classification. However, the analysis of GSL profiles is complex and requires dedicated technologies, that are currently not included in the commonly utilized high-throughput assays adopted in contexts such as molecular tumor boards.
In this study, we developed a method to discriminate the enzyme activity among the four series of the ganglioside metabolism pathway by incorporating transcriptome data and topological information of the metabolic network. We introduced three adjustment options for reaction activity scores (RAS) and demonstrated their application in both exploratory and comparative analyses by applying the method on neuroblastic tumors (NTs), encompassing neuroblastoma (NB), ganglioneuroblastoma (GNB), and ganglioneuroma (GN). Furthermore, we interpreted the results in the context of earlier published GSL measurements in the same tumors.
By adjusting RAS values using a weighting scheme based on network topology and transition probabilities (TPs), the individual series of ganglioside metabolism can be differentiated, enabling a refined analysis of the GSL profile in NT entities. Notably, the adjustment method we propose reveals the differential engagement of the ganglioside series between NB and GNB. Moreover, MYCN gene expression, a well-known prognostic marker in NTs, appears to correlate with the expression of therapeutically relevant gangliosides, such as GD2. Using unsupervised learning, we identified subclusters within NB based on altered GSL metabolism.
Our study demonstrates the utility of adjusting RAS values in discriminating ganglioside metabolism subtypes, highlighting the potential for identifying novel cancer subgroups based on sphingolipid profiles. These findings contribute to a better understanding of ganglioside dysregulation in NT and may have implications for stratification and targeted therapeutic strategies in these tumors and other tumor entities with a deregulated GSL metabolism.
糖鞘脂（GSLs）是由连接到聚糖部分的神经酰胺骨架组成的膜脂。神经节苷脂生物合成是GSL代谢的一部分，它涉及由特定酶催化的一系列反应，其中部分酶的底物特异性较差。GSLs在癌症中失调，因此可作为个性化治疗或亚型分类的潜在生物标志物。然而，GSL谱分析很复杂，需要专门技术，而这些技术目前未包含在分子肿瘤委员会等环境中常用的高通量检测中。
在本研究中，我们通过整合转录组数据和代谢网络的拓扑信息，开发了一种区分神经节苷脂代谢途径四个系列中酶活性的方法。我们引入了反应活性评分（RAS）的三种调整选项，并通过将该方法应用于神经母细胞瘤（NTs），包括神经母细胞瘤（NB）、神经节神经母细胞瘤（GNB）和神经节瘤（GN），展示了它们在探索性和比较性分析中的应用。此外，我们在先前发表的相同肿瘤的GSL测量结果背景下解释了结果。
通过使用基于网络拓扑和转移概率（TPs）的加权方案调整RAS值，可以区分神经节苷脂代谢的各个系列，从而能够对NTs实体中的GSL谱进行精细分析。值得注意的是，我们提出的调整方法揭示了NB和GNB之间神经节苷脂系列的不同参与情况。此外，MYCN基因表达是NTs中一个众所周知的预后标志物，似乎与治疗相关神经节苷脂如GD2的表达相关。使用无监督学习，我们基于改变的GSL代谢在NB中识别出亚群。
我们的研究证明了调整RAS值在区分神经节苷脂代谢亚型方面的实用性，突出了基于鞘脂谱识别新型癌症亚群的潜力。这些发现有助于更好地理解NTs中神经节苷脂的失调情况，并可能对这些肿瘤以及其他GSL代谢失调的肿瘤实体的分层和靶向治疗策略产生影响。
