Unique Role of Caffeine Compared to Other Methylxanthines (Theobromine, Theophylline, Pentoxifylline, Propentofylline) in Regulation of AD Relevant Genes in Neuroblastoma SH-SY5Y Wild Type Cells

Abstract

Methylxanthines are a group of substances derived from the purine base xanthine with a methyl group at the nitrogen on position 3 and different residues at the nitrogen on position 1 and 7. They are widely consumed in nutrition and used as pharmaceuticals. Here we investigate the transcriptional regulation of 83 genes linked to Alzheimer's disease in the presence of five methylxanthines, including the most prominent naturally occurring methylxanthines-caffeine, theophylline and theobromine-and the synthetic methylxanthines pentoxifylline and propentofylline. Methylxanthine-regulated genes were found in pathways involved in processes including oxidative stress, lipid homeostasis, signal transduction, transcriptional regulation, as well as pathways involved in neuronal function. Interestingly, multivariate analysis revealed different or inverse effects on gene regulation for caffeine compared to the other methylxanthines, which was further substantiated by multiple comparison analysis, pointing out a distinct role for caffeine in gene regulation. Our results not only underline the beneficial effects of methylxanthines in the regulation of genes in neuroblastoma wild-type cells linked to neurodegenerative diseases in general, but also demonstrate that individual methylxanthines like caffeine mediate unique or inverse expression patterns. This suggests that the replacement of single methylxanthines by others could result in unexpected effects, which could not be anticipated by the comparison to other substances in this substance class.

Keywords: caffeine; energy metabolism; lipid homeostasis; methylxanthines; oxidative stress; pentoxifylline; propentofylline; signal transduction; theobromine; theophylline.

Figure 1

Chemical structure of xanthine (left) and combinations of the side chains of its derivatives caffeine, theophylline, theobromine, pentoxifylline and propentofylline (right).

Figure 2

Design of the study. C: caffeine, TB: theobromine, TP: theophylline, P: pentoxifylline, PF: propentofylline.

Figure 3

Selection of housekeeping genes (HKGs) for normalization in SH-SY5Y cells. (A) Expression levels of the selected putative HKGs given as quantitative real-time RT-PCR cycle threshold (Cq) values in SH-SY5Y cells after treatment with solvent control, caffeine, theophylline, pentoxifylline, theobromine or propentofylline. The five most stable HKGs are presented in bold letters. (B) Expression stability values for 19 analyzed genes calculated by the NormFinder algorithm. Classification of stability values in most stable genes (0 < stability value < 0.37; green), intermediate genes (0.37 < stability value < 0.53; yellow) and least stable genes (0.53 < stability value < 0.69; red) was done in accordance with Penna et al. [25]. HKGs are ranked from the least to the most stable from left to right and the five most stable genes selected for normalization in SH-SY5Y cells (RN18S1, GPI, TOP1, YWHAZ and GAPDH) are highlighted by a box.

Figure 4

Transcriptional influence of caffeine, theobromine, theophylline, pentoxifylline and propentofylline in SH-SY5Y cells. (A) Transcriptional changes of genes related to the different pathways involved in oxidative stress, lipid and energy metabolism, signal transduction and gene expression, and Aβ- and tau-pathology and inflammation after treatment of human neuroblastoma cells (SH-SY5Y) for 24 h with 100 µM of the analyzed xanthine derivatives, in comparison to the solvent control, are illustrated in a heatmap. Fold changes greater than the standard deviation (yellow) are highlighted in red when downregulated and in green when upregulated. (B) The number of genes which are significantly changed by a single methylxanthine are summarized in a bar diagram. To calculate the significance of the observed effects we first performed an ANOVA for each gene to examine if there are differences in general. Afterwards, these p-values were adjusted with the false discovery rate method over all 83 analyzed genes. To determine which methylxanthine significantly changed the expression of a gene compared to the solvent control, we performed Dunnett’s test. To examine if there are significant differences in the distribution patterns of the significantly down- or upregulated genes influenced by the analyzed methylxanthines, we performed Fisher’s exact test.

Figure 5

Comparison of the transcriptional effects of caffeine, theobromine, theophylline, pentoxifylline and propentofylline. Results of the principal component analysis for each analyzed pathway are shown on the left. Fold changes and significances of those genes mediating the trends seen in the principal component analysis are illustrated in a bar diagram on the right for each pathway. Standard error of the mean is shown by the error bars and significance was calculated by Tukey’s honestly significant difference (HSD) post-hoc test and set as * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001.

Figure 6

Analysis of methylxanthine-mediated transcriptional effects, independent of cell type. The Venn diagram illustrates the genes of which the expressional changes, mediated by the analyzed methylxanthines (shown in different colors), are conserved over the three cell lines SH-SY5Y, HEPG2 and Calu-3. C: caffeine, TB: theobromine, TP: theophylline, P: pentoxifylline, PF: propentofylline.

Figure 7

Analysis of dose-dependent effects of methylxanthines. The expression of APH1b, ALS2, A2M, CASP4, ACHE and CASP4 are plotted against the concentrations of caffeine (A), theophylline (B), theobromine (C) and propentofylline (D) in diagrams, respectively. The Pearson correlation coefficient r and the corresponding p-value are shown for each gene.