Experimental Characterization Of Gene Regulatory Networks

SwitchGear Genomics

1455 Adams Drive # 2015
Menlo Park, CA, 94025
Website: http://www.switchgeargenomics.com

Experimental Characterization Of Gene Regulatory Networks

by Nathan D. Trinklein, Ph.D. and Shelley Force Aldred, Ph.D.

Introduction

Figure 1. Diagram of the SwitchGear Functional Promoter Macroarray. Each well of a microplate contains a unique reporter construct in a transfection ready format. After the addition of cells, the activity of each element can be measured while exposed to a compound or other stimulus.Click to enlarge.

The post-genome era has lead to great advances in high-throughput genomics studies. Genomic approaches such as microarray expression analyses and ChIP-chip transcription factor binding assays have provided a more comprehensive view of genetic pathways. These technologies provide valuable observational data but do not identify the functional connections within networks or explain the mechanism of gene regulation. Functional reporter assays provide an important additional layer of data for understanding mechanisms of regulation in genetic networks. Traditionally, reporter assays have been used on a very small scale to validate large-scale experiments like those described above. SwitchGear Genomics (Menlo Park, CA) has developed a novel approach that scales reporter assays to a much larger scale. The company builds and sells Functional Macroarrays that contain thousands of human promoters, UTRs and other regulatory elements encompassing many different disease-related pathways that are available as ready-to-use tools for cell-based studies. (Figure 1) In addition to Functional Macroarrays, SwitchGear has created an integrated platform that includes oligo microarrays for studying transcription factor binding and histone modifications and a novel assay for measuring patterns of DNA methylation on a genome-wide scale. Ultimately, this new technology and approach enables a detailed analysis of regulatory pathways and gene networks in living cells.

Results

Figure 2. Integrated measures of promoter activity, transcript levels, and transcription factor binding for 96 human genes. This graph represents three types of data gathered for 96 human genes in the HCT116 human liver cell line: 1) transcript levels of the endogenous human genes measured by quantitative RT-PCR are shown on the x axis, 2) binding of the TAF1 basal transcription factor as measured by chromatin IP is shown on the y axis, 3) functional promoter activity measured by a promoter-luciferase reporter experiment where the diameter of the bubble is equal to promoter activity. The genes labeled A are examples of genes with high promoter activity, high endogenous transcript levels, but no detectable TAF1 binding. The genes labeled B are examples of genes with medium promoter activity, high TAF1 binding, but low endogenous transcript levels.Click to enlarge.

An example of the power of combining multiple data types is shown in Figure 2. In this experiment, endogenous transcript levels, promoter activity, and basal transcriptional factor binding were all measured for 96 genes. There are a number of genes on the plot that fit with the prevailing model that binding of the TFIID basal transcription complex (as measured by TAF1 in this example) correlates with gene expression (as measured by promoter activity and endogenous transcript levels). However, there are exceptions to this pattern that are very interesting. The group of genes labeled A in Figure 1 that have high endogenous transcript levels and promoter activity but do not show evidence of basal transcription factor binding are genes that may be expressed by a TAF1-independent mechanism. Likewise, those genes with significant transcription factor binding and strong promoter activity but with low endogenous transcript levels are candidates for being post-transcriptionally regulated and have a high rate of transcript turnover. These observations highlight the importance of integrating multiple independent experimental results, and specifically, of high-throughput promoter assays in providing a complete picture of gene regulation.

To test the reproducibility, dynamic range, and inducible activity of our functional reporter assays, we transfected four independent preparations of a luciferase construct containing the promoter of the lactate dehydrogenase A gene (LDHA). LDHA is known to be induced by hypoxia due to activation by HIF1. The activity of the LDHA promoter was measured during exposure to DFO, a known activator of HIF1. Induction data are shown in Figure 3. The hypoxia induction experiment shows that functional promoter assays conducted in high-throughput format are also capable of measuring induction events over a time-course and show consistent results between independent construct preparations and over a broad dynamic range.

Furthermore, our high-throughput reporter platform can also be used to measure post-transcriptional regulation mediated through the 3’ UTR of a transcript. We used our platform to compare the transcriptional and post-transcriptional regulation of the transferring receptor (TFRC) gene during hypoxia by measuring promoter activity and UTR activity before and after exposure to DFO. The TFRC promoter was induced 2.82 fold upon treatment with DFO, and the TFRC UTR increased reporter activity 2.86 fold upon treatment with DFO (see Figure 4). The regulation of TFRC expression represents a case in which studying the function of both the transcriptional and post-transcriptional regulatory elements is indispensable for gaining a complete understanding of the gene’s regulation.

Conclusion

Figure 3. Activity of the LDHA promoter during a hypoxia-induction time-course. The activities of four independent preparations of a luciferase-based reporter construct driven by the LDHA promoter were measured across a time-course of DFO exposure. DFO is a chemical inducer of the hypoxia response. The experiment was conducted in 96-well format in HT1080 cells. The functional promoter assay yields consistent results across the four replicates and provides new insights into the kinetic behavior of the LDHA promoter in response to hypoxic conditions. This experiment is an example of how such time-course experiments may be conducted with SwitchGear tools in a high-throughput format that may easily be extended to include hundreds of promoters of interest. Click to enlarge.

The goal of SwitchGear Genomics is to enable groups to measure the function of thousands of regulatory elements in living cells under experimental conditions of interest. SwitchGear’s functional tools have successfully been used to identify new members of gene networks and have also been used as efficient and cost-effective screening tools to detail how compounds affect entire pathways rather than just a single marker. These new screening tools will also enable researchers to greatly enrich existing genomic datasets and provide functional annotation that will thoroughly describe the mechanism of action of a regulatory pathway. The data shown here highlight the ability to gather highly reproducible data over a broad dynamic range and the ability to scale this approach to hundreds or thousands of promoters in a single experiment.

About the authors

Nathan D. Trinklein is a co-founder of SwitchGear Genomics. As CEO and CSO, he manages all technology development, genome informatics, and applications in the area of regulatory network/pathway analysis. Shelley Force Aldred is a co-founder of SwitchGear Genomics. As COO and President, she directs all lab operations.

More information about the technology discussed in this article is available from:
SwitchGear Genomics
650-323-6763
www.switchgeargenomics.com

Methods

HT1080 cells (ATCC, Manassas, VA) were grown and seeded in Advanced MEM (Invitrogen Corp., Carlsbad, CA) with 10% FBS, 100 /ml penicillin, 100 g/ml streptomycin. The cells were seeded at 3750 cells/well in 100 l total volume or 1500 cells/well in 30 l total volume (96- and 384-well formats respectively) in white tissue culture treated multi-well plates and incubated for 24 hours prior to transfection at 37 C in 5% CO₂. For each 96-well transfection we mixed 1.67 l experimental plasmid (30 ng/l), 0.3 l Fugene-6 Transfection Reagent (Roche Applied Science, Indianapolis, IN), and 3.03 l OptiMEM (Invitrogen) and incubated for 30 minutes at room temperature before adding to the well of seeded cells. For 384-well transfections, we conducted the same procedure with 1 l plasmid (30 ng/l), 0.12 l Fugene-6, 1.88 l OptiMEM.

For the hypoxia experiment, we incubated the transfected cells for 8 hours and then added DFO to a final concentration of 100 M (Sigma-Aldrich, St. Louis, MO) and continued incubation for 3, 6, 9, 24 and 34 hours before conducting the assay for luciferase activity. Each transfection and assay experiment was conducted with four replicate wells.

Figure 4. Hypoxia-inducible activity of the TFRC promoter and 3’ UTR The activity of the TFRC promoter and 3’ UTR were separately measured by SwitchGear’s high-throughput functional macroarrays in HT1080 human fibrosarcoma cells. Relative reporter activity is shown on the y-axis before and after 24 hours of exposure to DFO.Click to enlarge.

For the TFRC experiment, we incubated cells in the presence of DFO for 24 hours. To assay luciferase activity in each well, we added 100 l or 30 l (96- and 384-well formats respectively) of Steady-Glo Luciferase Reagent (Promega Corp., Madison, WI) directly to the well of transfected cells and incubated for 15 minutes shaking at room temperature. We then used the LMax384 Luminometer (Molecular Devices, Inc., Sunnyvale, CA) to read the luminescence signal from each well for 2 seconds. For the hypoxia experiment, we calculated the fold induction in activity over that of the promoter at time-point zero.