Dietary effects of arachidonate-rich fungal oil and fish oil on murine hepatic and hippocampal gene expression

Abstract

Background: The functions, actions, and regulation of tissue metabolism affected by the consumption of long chain polyunsaturated fatty acids (LC-PUFA) from fish oil and other sources remain poorly understood; particularly how LC-PUFAs affect transcription of genes involved in regulating metabolism. In the present work, mice were fed diets containing fish oil rich in eicosapentaenoic acid and docosahexaenoic acid, fungal oil rich in arachidonic acid, or the combination of both. Liver and hippocampus tissue were then analyzed through a combined gene expression- and lipid- profiling strategy in order to annotate the molecular functions and targets of dietary LC-PUFA.

Results: Using microarray technology, 329 and 356 dietary regulated transcripts were identified in the liver and hippocampus, respectively. All genes selected as differentially expressed were grouped by expression patterns through a combined k-means/hierarchical clustering approach, and annotated using gene ontology classifications. In the liver, groups of genes were linked to the transcription factors PPARalpha, HNFalpha, and SREBP-1; transcription factors known to control lipid metabolism. The pattern of differentially regulated genes, further supported with quantitative lipid profiling, suggested that the experimental diets increased hepatic beta-oxidation and gluconeogenesis while decreasing fatty acid synthesis. Lastly, novel hippocampal gene changes were identified.

Conclusions: Examining the broad transcriptional effects of LC-PUFAs confirmed previously identified PUFA-mediated gene expression changes and identified novel gene targets. Gene expression profiling displayed a complex and diverse gene pattern underlying the biological response to dietary LC-PUFAs. The results of the studied dietary changes highlighted broad-spectrum effects on the major eukaryotic lipid metabolism transcription factors. Further focused studies, stemming from such transcriptomic data, will need to dissect the transcription factor signaling pathways to fully explain how fish oils and arachidonic acid achieve their specific effects on health.

Figure 1

Graphical representation of the trend and spread of the hepatic microarray data. Blue dots indicate the highest fold change for each probe set on the Affymetrix Mu11K chip, where the highest fold change was identified by examining all fold changes in every possible pairwise comparison between diets. Those probe sets that were identified as absent (by Affymetrix software) across all dietary treatments were treated as unchanged and therefore not considered in the limit fold change (LFC) selection model. The trend within this dataset indicates that as absolute expression increases (along the x-axis), fold changes decrease, i.e. the chance of seeing large fold changes with highly expressed genes is minimal. The 5% LFC line of best fit (red line) takes this observed trend into account and identifies the upper 5% of those probe sets with the greatest fold changes within predefined bins (model parameters fully explained in [27]). The 329 probe sets that lie above the 5% LFC line are those genes defined as differentially regulated. Green circles represent the hepatic genes discussed in the current manuscript and black squares represent those genes validated by real time PCR. When examining the spread of the genes discussed and validated, it is apparent that genes were selected across the entire range of absolute values and not concentrated in one area. Furthermore, as the validation data (which had 86% concordance with microarray data; see [27]) lies across the entire range of absolute expression levels, a high degree of confidence is established for the genes discussed within the present manuscript.

Figure 2

Venn diagrams depicting the top 10 GO functional categories in the hippocampus and liver. Examining the top 10 GO classifications for cellular component (CC), biological process (BP), and molecular function (MF) indicated that functional commonalties (overlapping circles) in response to LC-PUFAs exist between the two organs, implying that molecular responses to dietary LC-PUFA occurs via common biological pathways rather than through identical genes. Additional organ-specific functions are also indicated (individual circles).