Background

Metabolic processes within cells play a major role in tumor adaptation, growth, and proliferation [1]. In the case of glioblastoma (GB), our previous in vitro studies focused on tumor plasticity driven by the hypoxic and acidic microenvironment and highlighted the critical role of lactate dehydrogenases (LDHA, LDHB) in this metabolic symbiosis [2]. We employed omics technologies—specifically transcriptomics and metabolomics—because they provide complementary insights into metabolic pathways. Transcriptomics captures the initial stage of gene expression through the analysis of mRNA transcripts, while metabolomics sheds light on the enzymatic activities associated with transcript translation.

To further characterize metabolic dynamics, we used stable-isotope resolved metabolomics (SIRM), also referred to as tracer or labeled metabolomics [3]. SIRM involves the administration of a stable isotope–labeled substrate to the cells or organisms, allowing to trace the integration of these isotopes into metabolic pathways, thus revealing subsequent biochemical reactions [3–5]. While ¹³C is the most commonly used isotope, other isotopes are also employed. The resulting corrected data contain the total abundances and labeling information, such as isotopologues and mean enrichment [4,6]. However, previous bioinformatics solutions for analyzing these datasets have exhibited limitations, including statistical frameworks or accessibility [7–11]. Furthermore, flux estimation often requires substantial computational resources and complex parameterization [12,13], creating a barrier to its adoption.

This protocol begins with the preparation of glioblastoma cell lines. Two distinct cell lines are used, P3 sgCont and sgLDHA/B, both generated from P3 wild-type glioblastoma cells through CRISPR-Cas9 editing [2]. For simplicity, we refer to sgCont as control throughout this paper. The sgLDHA/B cell line corresponds to a double knockout (KO) of the LDHA and LDHB genes and is referred to as LDHAB-KO. Along these cell lines, we provide the corresponding SIRM and transcriptomics datasets.

The primary focus of this paper is the data analysis using our recently published tool, DIMet [6], available in its Galaxy interface. DIMet is the first comprehensive tool designed for the differential analysis of tracer metabolomics data integrated with transcriptomics data. This protocol provides a detailed step-by-step guide for using DIMet in Galaxy. Specifically, we performed differential analyses between conditions and/or time points to evaluate total metabolite abundances and isotope labeling. By interpreting these two levels of information together, it becomes possible to infer variations in the production and consumption rates of metabolites, providing a proxy for flux differences [4,9].

We further combined metabolomics and transcriptomics datasets into DIMet's pathway-based metabolograms, addressing the limitations of previous tools. DIMet offers several key advantages: (i) it accommodates the unique requirements of SIRM data, (ii) it integrates data in a pathway-centric manner, (iii) it generates visually ergonomic outputs, and (iv) all functionalities are accessible via the free, user-friendly Galaxy interface. While complex analyses involving multiple isotopes and substrates are not yet supported, DIMet’s features and accessibility more than compensate for this limitation. Without restrictions on the system, organism, or isotope-substrate combination, DIMet in Galaxy can be used for differential analysis of any corrected tracer metabolomics dataset.