Text mining of full-text journal articles combined with gene expression analysis reveals a relationship between sphingosine-1-phosphate and invasiveness of a glioblastoma cell line

Abstract

Background: Sphingosine 1-phosphate (S1P), a lysophospholipid, is involved in various cellular processes such as migration, proliferation, and survival. To date, the impact of S1P on human glioblastoma is not fully understood. Particularly, the concerted role played by matrix metalloproteinases (MMP) and S1P in aggressive tumor behavior and angiogenesis remains to be elucidated.

Results: To gain new insights in the effect of S1P on angiogenesis and invasion of this type of malignant tumor, we used microarrays to investigate the gene expression in glioblastoma as a response to S1P administration in vitro. We compared the expression profiles for the same cell lines under the influence of epidermal growth factor (EGF), an important growth factor. We found a set of 72 genes that are significantly differentially expressed as a unique response to S1P. Based on the result of mining full-text articles from 20 scientific journals in the field of cancer research published over a period of five years, we inferred gene-gene interaction networks for these 72 differentially expressed genes. Among the generated networks, we identified a particularly interesting one. It describes a cascading event, triggered by S1P, leading to the transactivation of MMP-9 via neuregulin-1 (NRG-1), vascular endothelial growth factor (VEGF), and the urokinase-type plasminogen activator (uPA). This interaction network has the potential to shed new light on our understanding of the role played by MMP-9 in invasive glioblastomas.

Conclusion: Automated extraction of information from biological literature promises to play an increasingly important role in biological knowledge discovery. This is particularly true for high-throughput approaches, such as microarrays, and for combining and integrating data from different sources. Text mining may hold the key to unraveling previously unknown relationships between biological entities and could develop into an indispensable instrument in the process of formulating novel and potentially promising hypotheses.

Figure 1

Frequency distribution of relationship types. Relationships were identified from sentence level extraction using LexiQuest Mine (SPSS, Chicago,IL) and patterns developed as previously described [25]. In total, 54 types of relationships were identified. The name and percentage occurrence of the top 10 relationships are shown. The most frequent type of relationship refers to 'interaction' (14.4%); the least frequent type refers to 'homo-oligomerization', featured in only two patterns.

Figure 2

Schematic diagram of the text mining analysis pipeline. Full-text articles are downloaded and processed using the download agent GetItRight. The resulting HTML files are converted to XML. Biological entities (genes, proteins) and their relationships (activation, inhibition, etc.) are extracted from LexiQuest Mine (SPSS, Chicago, IL). The resulting patterns are stored in the text mining data warehouse. The text mining data is matched with results from a differential gene expression experiment.

Figure 3

S1P-Network. Interaction network for differentially expressed genes (sentences related to S1P). The directed pseudograph of relationships related to S1P was generated as described in materials and methods. Seed vertices (shown in red) are the gene names from the list of 72 differentially expressed genes. Blue vertices and bold purple arcs represent genes and relationships that were found in this interaction network and in the interaction network related to invasivity shown in Figure 4.

Figure 4

nvasion-Network. Gene interaction network of genes related to invasivity. The directed pseudograph of relationships related to invasivity was generated as described in materials and methods. Seed vertices (shown in red) are the gene names from the list of 72 differentially expressed genes. Blue vertices and bold purple arcs represent genes and relationships that were found in this interaction network and in the interaction network related to S1P shown in Figure 3. Genes directly related to matrix metalloproteinases (key components of invasivity) are highlighted by the mustard-colored ovals.

Figure 5

Intersection-Network. Gene interaction network derived from an intersection of the S1P- and invasion-network. This interaction network was hand drawn using gene names found in the S1P (Figure 3) and invasivity (Figure 4) networks as input vertices. In addition to the direction of the relationship shown by the arrow, the type is also depicted as text superimposed onto the arrow. The resulting graph contains several genes differentially expressed in the presence of S1P. These are shown in the red ovals. Genes directly related to matrix metalloproteinases are highlighted by the mustard-colored ovals. Key relationships describing the most direct connections between S1P and invasivity are highlighted by the bold purple arrows. The red arrow indicates that S1P induced transcription of NRG-1 in the microarray experiments.

Figure 6

Example of transitive dependencies.

Figure 7

Example of interaction network.

Figure 8

Pruning strategy for network construction.