Search

Hepatic lipid deposition and inflammation represent risk factors for hepatocellular carcinoma (HCC). The mRNA-binding protein tristetraprolin (TTP, gene name ZFP36) has been suggested as a tumor suppressor in several malignancies, but it increases insulin resistance. The aim of this study was to elucidate the role of TTP in hepatocarcinogenesis and HCC progression. Employing liver-specific TTP-knockout (lsTtp-KO) mice in the diethylnitrosamine (DEN) hepatocarcinogenesis model, we observed a significantly reduced tumor burden compared to wild-type animals. Upon short-term DEN treatment, modelling early inflammatory processes in hepatocarcinogenesis, lsTtp-KO mice exhibited a reduced monocyte/macrophage ratio as compared to wild-type mice. While short-term DEN strongly induced an abundance of saturated and poly-unsaturated hepatic fatty acids, lsTtp-KO mice did not show these changes. These findings suggested anti-carcinogenic actions of TTP deletion due to effects on inflammation and metabolism. Interestingly, though, investigating effects of TTP on different hallmarks of cancer suggested tumor-suppressing actions: TTP inhibited proliferation, attenuated migration, and slightly increased chemosensitivity. In line with a tumor-suppressing activity, we observed a reduced expression of several oncogenes in TTP-overexpressing cells. Accordingly, ZFP36 expression was downregulated in tumor tissues in three large human data sets. Taken together, this study suggests that hepatocytic TTP promotes hepatocarcinogenesis, while it shows tumor-suppressive actions during hepatic tumor progression.

The authors regret that in the published article the repetitive display of some values was not precisely indicated. The authors apologize to the readers of Nanomedicine for any inconvenience or confusion this may have caused. The semi-automated and conceptual array-based design presented in the study allowed us to conduct and analyze all displayed transfection experiments in parallel as described in the Supplemental Materials & Methods section. Accordingly, the same control and reference values were re-used in different analysis steps. To our concern, we unintentionally displayed some values repetitively without explicitly describing this fact. Although our mistakes do not alter the result and conclusion which can be drawn from our study, we would like to account for the lack of precision and clarify: 1. The cell viability, GADPH and TFR results for the negative control NC as displayed in Figure 2, A and C were again displayed in part in Figure 4 and in Supplemental Figure 4. In all cases the value is set to 1 as reference point.2. Cell viability results of the negative control NC and the positive controls PC1 and PC2 shown in Figure 2, A and C were re-displayed in Figure 5, A and B for easy comparison.3. Ki67 knockdown results displayed in Figure 5, A and B are partly included in Supplemental Figure 4 and TFR knockdown results in Figure 4 were re-used in Supplemental Figure 5.4. The sentences “Antibodies targeting TFR, GAPDH and Ki67 were applied to RPPAs carrying lysates of siKi67 transfected MCF7 and MCF12A.” in the figure legend of Supplemental Figure 4 should be corrected to “Antibodies targeting TFR, GAPDH and Ki67 were applied to RPPAs carrying lysates of siKi67 transfected MCF12A (A) and MCF7 (B).”.

Identifying functional modules or novel active pathways, recently termed de novo pathway enrichment, is a computational systems biology challenge that has gained much attention during the last decade. Given a large biological interaction network, KeyPathwayMiner extracts connected subnetworks that are enriched for differentially active entities from a series of molecular profiles encoded as binary indicator matrices. Since interaction networks constantly evolve, an important question is how robust the extracted results are when the network is modified. We enable users to study this effect through several network perturbation techniques and over a range of perturbation degrees. In addition, users may now provide a gold-standard set to determine how enriched extracted pathways are with relevant genes compared to randomized versions of the original network.

High-throughput screening is extensively applied for identification of drug targets and drug discovery and recently it found entry into toxicity testing. Reverse phase protein arrays (RPPAs) are used widespread for quantification of protein markers. We reasoned that RPPAs also can be utilized beneficially in automated high-throughput toxicity testing. An advantage of using RPPAs is that, in addition to the baseline toxicity readout, they allow testing of multiple markers of toxicity, such as inflammatory responses, which do not necessarily cumulate in cell death. We used transfection of siRNAs with known killing effects as a model system to demonstrate that RPPA-based protein quantification can serve as substitute readout of cell viability, hereby reliably reflecting toxicity. In terms of automation, cell exposure, protein harvest, serial dilution and sample reformatting were performed using a robotic screening platform. Furthermore, we automated sample tracking and data analysis by developing a bundled bioinformatics tool named “MIRACLE”. Automation and RPPA-based viability/toxicity readouts enable rapid testing of large sample numbers, while granting the possibility for flexible consecutive quantification of additional user-defined toxicity markers.

High-throughput screening of genome wide siRNA- or compound libraries is currently applied for drug target and drug discovery. Commonly, these approaches deal with sample numbers ranging from 100,000 to several millions. Efforts to decrease costs and to increase information gained include multiplexing readouts, but this has a natural limitation. High-content screening via image acquisition and analysis allows multiplexing of few parameters, but is connected to substantial time consumption and complex logistics. We report on integration of Reverse Phase Protein Arrays (RPPA)-based readouts into automated robotic high-throughput screens, which allows subsequent protein quantification. In this integrated solution, samples are directly forwarded to automated cell lysate preparation and preparation of dilution series, including reformatting to a protein spotter-compatible format after the high-throughput screening. Tracking of huge sample numbers and data analysis from a high-content screen to RPPAs is accomplished via MIRACLE, a custom made software suite developed by us. To this end, we demonstrate that the RPPAs generated in this manner deliver reliable protein readouts and that GAPDH and TFR levels can be used for normalization of protein amounts as they correlate highly with measured cell viability. Thus, this strategy offers ample options for multiplexed screen readouts via consecutive protein quantifications.

Selecting the most promising treatment strategy for breast cancer crucially depends on determining the correct subtype. In recent years, gene expression profiling has been investigated as an alternative to histochemical methods. Since databases like TCGA provide easy and unrestricted access to gene expression data for hundreds of patients, the challenge is to extract a minimal optimal set of genes with good prognostic properties from a large bulk of genes making a moderate contribution to classification. Several studies have successfully applied machine learning algorithms to solve this so-called gene selection problem. However, more diverse data from other OMICS technologies are available, including methylation. We hypothesize that combining methylation and gene expression data could already lead to a largely improved classification model, since the resulting model will reflect differences not only on the transcriptomic, but also on an epigenetic level. We compared so-called random forest derived classification models based on gene expression and methylation data alone, to a model based on the combined features and to a model based on the gold standard PAM50. We obtained bootstrap errors of 10-20% and classification error of 1-50%, depending on breast cancer subtype and model. The gene expression model was clearly superior to the methylation model, which was also reflected in the combined model, which mainly selected features from gene expression data. However, the methylation model was able to identify unique features not considered as relevant by the gene expression model, which might provide deeper insights into breast cancer subtype differentiation on an epigenetic level. Selecting the most promising treatment strategy for breast cancer crucially depends on determining the correct subtype. In recent years, gene expression profiling has been investigated as an alternative to histochemical methods. Since databases like TCGA provide easy and unrestricted access to gene expression data for hundreds of patients, the challenge is to extract a minimal optimal set of genes with good prognostic properties from a large bulk of genes making a moderate contribution to classification. Several studies have successfully applied machine learning algorithms to solve this so-called gene selection problem. However, more diverse data from other OMICS technologies are available, including methylation. We hypothesize that combining methylation and gene expression data could already lead to a largely improved classification model, since the resulting model will reflect differences not only on the transcriptomic, but also on an epigenetic level. We compared so-called random forest derived classification models based on gene expression and methylation data alone, to a model based on the combined features and to a model based on the gold standard PAM50. We obtained bootstrap errors of 10-20% and classification error of 1-50%, depending on breast cancer subtype and model. The gene expression model was clearly superior to the methylation model, which was also reflected in the combined model, which mainly selected features from gene expression data. However, the methylation model was able to identify unique features not considered as relevant by the gene expression model, which might provide deeper insights into breast cancer subtype differentiation on an epigenetic level.

Short interfering RNAs (siRNAs) are an attractive option for the next generation of anti-cancer drugs, due to their ability to selectively inactivate each of the 20,000 human genes. Suitable siRNAs for the development of such drugs are commonly recovered via automated genome-wide siRNA screens. In a differential screen, comparing e.g. cancer cells versus normal cells, up to 120,000 cell experiments are performed in a highly parallelized fashion. The readout options of such high-throughput screens are, however, generally limited to one or very few parameters.We report on establishing a Reverse Phase Protein Array (RPPA)-based readout format integrated into robotic siRNA screening. This technique would allow post-screening high-throughput quantification of protein changes. Recently, breast cancer stem cells (BCSCs) have attracted much attention, as a tumor- and metastasis-driving subpopulation, and thus this methodology was laid out for and tested using mammospheres which are believed to be enriched for BCSCs.