Project 1: The functional genomics of radiation bystander responses

Project Leader: Sally A. Amundson

Overiew
While traditional studies of cellular ionizing radiation responses have focused on the direct deposition of energy in the nucleus, it is now recognized that cells also respond to extra-nuclear radiation damage, and even to extra-cellular radiation damage via the bystander effect. However, the differential signal transduction pathways regulating the responses to damage in different cellular compartments have not been well elucidated. This project seeks to harness the power of microarray profiling and functional genomics in conjunction with the single-cell / single-particle microbeam irradiator in order to gain insight into the mechanisms of signaling between cellular compartments and between cells in response to radiation damage.

Research Aims

This project encompasses four broad aims:

1. To develop gene expression profiles from cytoplasmic vs. nuclear irradiation in primary human skin fibroblasts and normal human bronchial epithelial cells using the single-cell / single-particle microbeam to deliver defined numbers of alpha-particles to cell nuclei or cytoplasm, and to investigate the signal transduction pathways involved.
2. To develop gene expression profiles of cultured normal human fibroblasts and epithelial cells responding to bystander signals induced by -particles delivered both as broad field and microbeam irradiation. Key components of the pathways implicated in the microarray studies will be interrupted with siRNA or chemical inhibitors to ablate the bystander response.
3. Identify targets of the arachidonic acid signaling cascade by gene expression profiling. Signaling through the arachidonic acid cascade will be interrupted by a COX2-specific chemical inhibitor. Compare global RNA expression in the presence or absence of functional PTGS2 (aka COX2) with and without direct or bystander irradiation.
4. Develop gene expression profiles of bystander radiation responses in artificial 3-D human tissue models (skin and bronchial epithelium) more closely representing normal tissue micro-architecture and micro-environments.

Research Highlights

Whole genome profiling of bystander response

We have used the RARAF track segment facility and the “Outer/Inner” strip dish approach to compare gene expression in normal human lung fibroblasts from Cambrex, IMR90, and AG1522 cells after direct alpha-particle or bystander irradiation. Agilent whole genome microarrays were used to profile gene expression at times from 0.5 to 24 hours after exposure. We scanned the microarrays using the Agilent DNA Microarray Scanner with Feature Extract software for data extraction, background subtraction and normalization, followed by analysis with BRB ArrayTools. Downstream analyses used Panther and DAVID for gene ontology analyses, and Ingenuity Pathway Analysis (IPA) for network analysis.

Aanalysis of the IMR90 data indicates a large initial gene expression response, tapering off by 24 hours after exposure. Analysis of the entire time-course is ongoing, and includes collaborative efforts with members of the Biostatistics Department to develop new statistical methodologies for interpreting such microarray time series data, as currently available tools are not adequate. As the largest response in bystanders occurred at the earliest time measured, we are also pursuing these early genes in greater detail.

Our initial analysis focused on genes that were differentially expressed 4 hr after treatment in bystander cells and the resulting ontology and pathway analysis indicated significant regulation of gene expression in irradiated and bystander cells by the transcription factor NFkappaB.

(Figure from Ghandhi et al., 2008 BMC Med Genomics 1: 63.) Ingenuity Pathways Analysis was used to extract nodes having direct regulatory interactions with p53 or NFκB from the set of genes significantly responding to radiation. Expression levels are overlaid on the nodes so that up-regulated genes are displayed in red and down-regulated genes in green. The scale bar shows the range of fold-changes. Panel A) p53 regulated genes are overlaid with relative expression levels in irradiated (left) or bystander (right) cells 4 hours after treatment. Just under half the genes changing in irradiated cells respond in bystanders. Panel B) NF κB regulated genes overlaid with relative expression in irradiated (left) and bystander (right). The responses are nearly identical at the 4-hour time point.

We have also investigated the expression of specific genes that are NFkappaB targets in bystanders at early time points and found that some of these genes such as IL-8, PTGS2/COX-2 and BCL2A1 show a coordinated biphasic induction pattern.

(Figure from Ghandhi et al., 2008 BMC Med Genomics 1: 63.) Quantitative real-time RT-PCR was used to monitor expression of A) CDKN1A; B) PTGS2; C) BCL2A1 and D) IL8 at 0.5, 1, 2, 4, 6 and 24 hours after direct irradiation (open circles) or bystander exposure (black squares) of IMR-90 cells. Gene expression was normalized to ACTB mRNA levels and is relative to expression in time-matched controls (dashed line). Points are the mean and standard error of four independent experiments.

3D tissues for bystander irradiation and gene expression

We have used the EPI-200 EpiDerm^TM system from MatTek as a 3-dimensional tissue model. Our 3D tissue studies have benefited from the design and manufacture by Gary Johnson in the Technical Core of new tissue holders with shielding masks that allow our bystander studies in 3D tissues to be performed using the track-segment irradiation mode. The mask shields most of the tissue, while exposing only a narrow strip of tissue to direct irradiation (See figure). The major advantage of this approach is the maintenance of a more precise registration with the plane of irradiation, allowing greater precision throughout the subsequent assays. A device for performing the initial slicing of the tissues in situ (in their holders) was also built for this project, and further enhanced precision.

After our initial success in obtaining the high-quality RNA needed for microarray studies from sliced, unfixed 3D tissues, we have completed a series of experiments comparing gene expression of these 3D tissues to direct irradiation with alpha-particles and x-rays. Based on our preliminary studies with this model, which indicated a delayed gene expression response to radiation compared to that typically seen in 2D cell culture, these studies utilized a 24-hour incubation after irradiation. The results of this analysis were quite impressive, with very good agreement between the repeated samples. The close reproducibility of the controls indicates that background variability between individual tissues or between different batches of tissues is minimal, and will not complicate interpretation of the bystander responses. Analysis with BRB Array Tools revealed more than 500 genes that were differentially expressed at the p<0.001 level following 50 cGy alpha-particles, with a false discovery rate less than 2.5%. A consensus set of 83 genes was identified that responded similarly to all irradiation conditions tested. The similarity of these responses in different tissue samples can be clearly seen in the Multi-Dimensional Scaling (MDS) plot in the accompanying figure.

Comparison	# p<0.001	FDR
X-ray vs. Control	388	<4%
Alpha vs. Control	537	<2.5%
Common to alpha + 8Gy x-ray	451	<3%

Number of genes significantly responding to irradiation conditions in EPI-200 tissues.

Key elements of the bystander response, such as regulation of PTGS2, appear to be retained in the 3-Dimensional tissues, and a detailed analysis of gene expression as a function of distance from the irradiated plane of tissue is in progress. Genes upregulated in bystander tissue, such as DMBT1, and down-regulated, such as MMP1, and the gene expression effect seems to fall off between 750 and 100 µm from the plane of irradiation. The timing of the response also appears to differ somewhat from that seen in the 2-dimensional fibroblast cultures.