NOTE: NCE to 5/29/2009 has been approved, per A. Chu/ARC (10/08)

NOTE: NCE to 5/31/2009, from 11/30/2008, is in process, per A. Chu/ARC (9/08)

The intent of this 5-year project NSCOR at Lawrence Berkeley National Laboratory Life Sciences Division is to provide a comprehensive examination of HZE effects from the initial damage, to early cellular HMEC responses, to persistent functional precursors of carcinogenesis. We have addressed the mechanisms, nature, and frequency of DNA damaging events; mechanisms of DNA repair and misrepair, early signal transduction mechanisms; immediate and long-term, reversible and irreversible gene expression changes; cellular remodeling and reorganization; potential mechanisms of tissue "repair" and matrix effects; and the cellular and molecular mechanisms of HZE/proton induced progression to a neoplastic phenotype.

In the context of the NSCOR, we have developed image analysis programs to quantify DNA damage response (e.g. radiation-induced foci) and specific organelles (i.e. centrosomes) using high-throughput imaging bioinformatics.

The characterization of non-malignant human mammary epithelial cells responses to ionizing radiation provides mechanistic insights to the processes relevant to human health risks from diagnostic and therapeutic radiation exposures.

The NSCOR has supported the development of a new media that enables human mammary epithelial cells to be maintained longer in culture.

Rationale: Epithelial cells depend on signals from the microenvironment to establish the requisite polarity for functional differentiation. Because most adult human solid tumors are epithelial in origin and because radiation effects are dictated by cellular genotype and phenotype, studies of HZE particle radiation biology using a physiologically relevant model will be the most informative regarding risk. The LBNL NSCOR developed methods for culturing human mammary epithelial cells (HMEC) derived from normal tissue as both monolayer and multicellular aggregates to study phenotypes and pathways that may be uniquely human, epithelial or multicellular. We examined how controlled cellular and microenvironment conditions affect early responses to HZE particle radiation-induced DNA damage in terms of repair, foci formation and gene expression, genomic instability, and persistent radiation-induced phenotypes and pathways that precede neoplasia. These surrogate endpoints were used to evaluate the neoplastic potential as a function of radiation quality, energy and dose.

Aim 1. We hypothesized that HZE particles such as 1GeV/u Fe-ions induce more persistent DNA damage than sparsely ionizing gamma rays (Rydberg et al., 2005). Studies using micronuclei supporting this hypothesis were published in Radiation Research (Groesser et al., 2007). To further test this idea, gamma-H2AX and 53BP1 radiation-induced foci (RIF) were measured in proliferating HMEC after exposure to low doses of sparsely and densely ionizing radiation. Preliminary results suggest that radiation induced gamma-H2AX and 53BP1 RIF in HMEC have different resolution dynamics as a function of radiation quality. While RIF numbers after gamma-ray irradiation returned to control level after about 12h, Fe-ion induced foci were still present at 24h. However, persistent RIF following HZE particles were eventually removed 24 – 48h after irradiation under the conditions used. The slower disappearance of foci after Fe-ion exposure could be due to the fact that HZE particles induce more complex DNA damage in the core of the ion track, which will be tested in future experiments using different ions and LET. We speculated that the persistent gamma-H2AX foci in Fe-ion irradiated non-proliferating cells could be due to the fact that not all DSB repair pathways for rejoining complex DSBs are available to the cell in G0/G1-phase or that damaged cells in a cycling population could be eliminated more efficiently via an apoptotic pathway. Apoptosis measurements in proliferating and non-proliferating MCF10A cells after exposure to different radiation qualities indicate that the apoptotic rate is very low in these cells and removal of damaged cells by apoptosis can only have a minor effect on our repair kinetic data. Another possibility is that the remaining gamma-H2AX foci represent chromatin remodeling rather than DSBs. These studies being prepared for publication (Groesser et al. in preparation).

Since energy is randomly deposited along high-LET particle paths, RIF along these paths should also be randomly distributed. A novel imaging approach shows that RIF were located preferentially at the interface between high and low DNA density regions, and were more frequent than predicted in regions with lower DNA density. The same preferential nuclear location was also measured for RIF induced by 1 Gy of low-LET radiation. This deviation from random behavior was evident only 5 min after irradiation for phosphorylated ATM RIF, while gamma H2AX and 53BP1 RIF showed more pronounced deviations up to 30 min after exposure. These data are published in PLoS Computational Biology (Costes et al., 2007). Following this high content analysis on fixed specimen showing that radiation induced foci (RIF) predominantly form in the euchromatin and at the interface between euchromatin/heterochromatin, we are using live cell imaging in cell line stably transfected with 53BP1-GFP from collaboration with Dr. Asaithamby Aroumougame from UT Southwestern NSCOR. The first movies obtained during NSRL07C and NSRL08A indicate two different types of foci movements. Many foci appear to move in a diffusive manner (Brownian motion), where as others show active movement with velocities that can reach 2 microns/min (Costes et al. in preparation).

We published one of the first comparisons of foci from gamma vs HZE irradiated cells (Costes et al., 2006). We have measured imaging properties of foci on and off tracks (e.g. foci frequency, foci size, foci intensity, and spatial organization of foci). Presumably, foci outside tracks are generated by delta-rays, where as foci along tracks are generated by the high-LET component of HZE. The conclusion of this analysis is a clear difference in foci formation kinetic for these two components. As previously reported, the maximum number of low-LET foci form over the course of 30 min to 1 hour. In contrast, foci along tracks appear within 2 min post-IR with a maximum frequency of approximately 0.7 foci/micron at these early times. Initial foci size and intensity are similar for both types of foci but low-LET foci have a slight size increase reaching a maximum 10 min post-radiation, high-LET foci seem to have a three-fold size increase over the first 30 min post-IR (Costes et al. in preparation).

Aim 2. While to date, many studies of ionizing radiation have focused on its effects on primary DNA structure, comparatively few have focused on its effects on gene methylation or phenotype. Such epigenetic phenomena have been increasingly implicated in carcinogenesis in human cells and tissues, and in some cases, may represent predominant mechanisms of oncogenesis. HMEC cultures spontaneously yield at low frequencies (1x10-5 to 1x10-8) variant p16(-) cell populations that are capable of long term growth (50-100 total population doublings), and which are susceptible to genomic instability associated with telomere dysfunction. We determined that 2 Gy X-irradiation causes increases in the appearance of p16(-) cells with long term growth potential. Flow cytometry confirmed that the differences in BrdU incorporation and senescence associated beta-galactosidase expression in the treated and untreated populations were statistically significant. Both sparsely and densely ionizing irradiation increased the frequency and rate with which HMEC escaped p16-associated stasis and acquired long-term growth potential inindependent experiments with HMEC from four individuals. We are currently preparing these data for publication and exploring the mechanistic underpinnings of this phenomenon (Mukhopadhyay et al. in preparation).

Our studies reporting that disrupted morphogenesis is a result of irradiated cells undergoing transforming growth factor ß 1 (TGFß) mediated epithelial to mesenchymal transition (EMT) were published in Cancer Research (Andarawewa et al., 2007). TGFß has many functions that orchestrate the response to DNA damage (Andarawewa et al., 2007). We compared the EMT induced by eqi-toxic doses of 1Gy of 1 GeV/amu 56Fe to 2Gy of 137Cs gamma radiation in both monolayer and 3D. Both exposures resulted in EMT in response to TGFß in monolayer and disrupted 3D morphogenesis when embedded in Matrigel. Radiation dose responses show that this effect has a very low threshold in that a single exposure of 3-100 cGy gamma radiation elicits the ‘same’ phenotypic switch, consistent with a non-targeted effect. TGF-ß induced EMT in irradiated HMEC is mediated by MAPK pathway, which crosstalks with integrins via reactive oxygen species (ROS). To test whether ROS could play a role in the EMT phenotype, we measured cellular ROS levels in irradiated, TGFß treated cells. We observed that double treated HMEC (184v and MCF10A) have sustained generation of ROS and increased mitochondria compared to single treatments (Andarawewa et al. in preparation).

Aim 3. We showed that exposure of finite lifespan HMEC in growth arrested monolayer culture to x-rays or Fe ions elicited severe karyotypic instability, defined as at least 3 subpopulations of cells per colony, in the clonal descendants of ~20% of clones. This phenotype was not seen in any of the unirradiated colonies. Although severe karyotypic instability was seen in colonies arising after exposure to the lowest dose of Fe ions (0.5 Gy), there was no evidence of a dose response and it was not evident after exposure to <2 Gy of x-rays. These data are now published in Radiation Research (Sudo et al., 2008). A dose response was seen for centrosome aberrations (CA) defined as 3 or more centrosomes/cell for the progeny of cells exposed in monolayers to x-rays or Fe ions. Unirradiated clones had a baseline frequency of 0.92% CA. Although there was no evidence for an RBE>1 for the induction of CA following irradiation in monolayer, irradiation of the same finite lifespan HMEC as growth arrested acinar structures in Matrigel increased colony forming ability of the descendents of the irradiated cells following exposure to x-rays, but eliminated the induction of centrosome aberrations. In contrast, exposure of these HMEC as preformed acinar structures provided no protection against Fe ion-induced cell killing or against the induction of centrosome aberrations suggesting that there may be an increased RBE for each of these endpoints in cells exposed to Fe ions in a more tissue-like architecture.

We have now published in Cancer Research that CA actually precede radiation-induced genomic instability in HMEC (Maxwell et al., 2008). Although CA induction is a linear dose response from 10-200 cGy, evidence of genomic instability in clonal assays showed a threshold between 10 and 50 cGy, suggesting another level of control on their persistence. We demonstrate that TGFß is a key mediator of genomic surveillance in human epithelial cells. TGFß did not prevent CA but rather deletes genomically unstable cells by p53 dependent apoptosis. Thus the targeted effect on centrosomes is countered by non-targeted signaling that limits the burden of potentially neoplastic cells. This model is now being extended to HZE exposures to test whether the differences seen in 3D as a function of radiation quality is based on induction efficiency or deletion efficiency.

Aim 4. We postulated that carcinogenesis can be thought as an emerging phenomenon arising from the perturbation of an organ’s homeostasis (Barcellos-Hoff and Costes, 2006). Emergent phenomena typically result from the interactions of individual entities (Barcellos-Hoff and Costes, 2006). By definition, a system with emerging properties cannot be reduced to its individual entities and as such “the whole is greater than the sum of its parts”. To address this concept, we began developing agent-based models (ABM), which are computer programs that represent a system as a collection of autonomous decision-making entities called agents. Each agent individually assesses its situation and makes decisions on the basis of a set of contextual rules. In the case of representing cells within an organism, agents may execute various behaviors appropriate for the system such as proliferation, differentiation or death. Such models have proven to be very useful in predicting emerging properties from complex systems. Using an agent-based model, we could predict growth population of post-stasis HMEC by simply assuming a 26 hour doubling time and their response to radiation as a function of clonogenic death that accelerates the emergence of variants. New studies using such approach are underway with very promising results. These studies complement those underway in the Tufts NSCOR and will provide a means of integrating biological data into a meaningful format for risk modeling.

In support of the overarching use of cell biology tools in radiation quality studies, we developed the BioSig imaging bioinformatics framework (Parvin et al., 2005). Information can be obtained from BioSig through either ad hoc queries or query by example. A previously developed 3D segmentation method quantitative analysis for 3D cell culture models (Raman et al., 2007) has been improved for its accuracy through multi-class labeling and geometric constraints (Parvin et al., 2007). The end result is that foci can be quantified in full 3D from volumetric representation. Once the data is analyzed, rendered surfaces are generated, and stored in the database. The analysis package has been designed to export the same feature-based representation in terms of morphology, structure, and localization. The schema supports more than 400 inter-related computed representations on a cell-by-cell basis. One recent novelty has been integration of computational and informatics content directly within the database (https://icbp.lbl.gov/biosig/home.do ). The net result has been direct visualization of user-selectable features through the BioSig imaging bioinformatics system. As a result, multi-dimensional representations can be visualized through heatmaps for hypothesis generation and direct visualization (Han et al., 2008).

Summary

The NSCOR data demonstrate that radiation quality affects types and persistence of DNA damage, genomic instability, cell fate, the microenvironment and responses to soluble signals. Potential interactions among these effects likely determine the probability of progressing to cancer ((Barcellos-Hoff, 2007). The LBNL NSCOR studies have shown that in general high vs low LET radiation is more effective in eliciting many biological processes like GIN and EMT that promote cancer development. However the fact that such responses do not exhibit typical dose dependence prevents the use of classic RBE comparisons. We propose that these studies motivate further examination of whether individual targeted and non-targeted phenomena (e.g. damage response, phenotype, genomic instability) interact in a unique manner to drive neoplastic transformation following high LET exposure.

NOTE: NCE to 5/29/2009 has been approved, per A. Chu/ARC (10/08)

The intent of this 5-year project NSCOR at Lawrence Berkeley National Laboratory Life Sciences Division is to provide a comprehensive examination of HZE effects from the initial damage, to early cellular HMEC responses, to persistent functional precursors of carcinogenesis. We have addressed the mechanisms, nature, and frequency of DNA damaging events; mechanisms of DNA repair and misrepair, early signal transduction mechanisms; immediate and long-term, reversible and irreversible gene expression changes; cellular remodeling and reorganization; potential mechanisms of tissue "repair" and matrix effects; and the cellular and molecular mechanisms of HZE/proton induced progression to a neoplastic phenotype.

In the context of the NSCOR, we have developed image analysis programs to quantify DNA damage response (e.g. radiation-induced foci) and specific organelles (i.e. centrosomes) using high-throughput imaging bioinformatics.

The characterization of non-malignant human mammary epithelial cells responses to ionizing radiation provides mechanistic insights to the processes relevant to human health risks from diagnostic and therapeutic radiation exposures.

The NSCOR has supported the development of a new media that enables human mammary epithelial cells to be maintained longer in culture.

We hypothesized that HZE particles such as 1GeV/amu Fe-ions induce more persistent DNA damage than sparsely ionizing gamma-rays. To address this, we measured gamma-H2AX and 53BP1 radiation-induced foci (RIF) in proliferating HMEC after exposure to low doses of sparsely and densely ionizing radiation. Preliminary results revealed the disappearance of radiation induced gamma-H2AX and 53BP1 RIF in HMEC have different dynamics as a function of radiation quality. While RIF numbers after gamma-ray irradiation returned to control level after about 12h, Fe-ion induced foci were still present at 24h. However, these persistent RIF were removed during the 24 – 48h time interval after irradiation under the conditions used. The slower disappearance of foci after Fe-ion exposure could be due to the fact that HZE particles induce more complex DNA damage in the core of the ion track, which will be tested in future experiments using different ions and LETs.

Since energy is randomly deposited along high-LET particle paths, RIF along these paths should also be randomly distributed. The probability to induce DSB can be derived from DNA fragment data measured experimentally by pulsed-field gel electrophoresis. We used this probability in Monte Carlo simulations to predict DSB locations in synthetic nuclei geometrically described by a complete set of human chromosomes, taking into account microscope optics from real experiments. As expected, simulations produced DNA-weighted random (Poisson) distributions. In contrast, the distributions of RIF obtained as early as 5 min after exposure to high LET (1 GeV/amu Fe) were non-random. This deviation from the expected DNA-weighted random pattern can be further characterized by “relative DNA image measurements”. This novel imaging approach shows that RIF were located preferentially at the interface between high and low DNA density regions, and were more frequent than predicted in regions with lower DNA density. The same preferential nuclear location was also measured for RIF induced by 1 Gy of low-LET radiation. This deviation from random behavior was evident only 5 min after irradiation for phosphorylated ATM RIF, while gammaH2AX and 53BP1 RIF showed more pronounced deviations up to 30 min after exposure. These data are now in press in PLoS Computational Biology.

While to date, many studies of ionizing radiation have focused on its effects on primary DNA structure, comparatively few have focused on its effects on epigenetic phenomena such as gene methylation or phenotype. Such epigenetic phenomena have been increasingly implicated in carcinogenesis in human cells and tissues, and in some cases, may represent predominant mechanisms of oncogenesis. In serum-free growth medium, HMEC from histologically normal breast tissue growth arrest after 5-20 population doublings, exhibiting senescent morphologies and elevated expression of p16. Epigenetic silencing of the cyclin-dependent kinase inhibitor p16INK4a (p16) in human cancers is often associated with methylation of the p16 promoter by unknown mechanism(s). Such HMEC cultures spontaneously yield at low frequencies (1x10-5 to 1x10-8) variant p16(-) cell populations that are capable of long term growth (50-100 total population doublings), and which are susceptible to genomic instability associated with telomere dysfunction. We sought to determine whether radiation would alter p16 expression, p16 promoter methylation, and/or long term growth potential in HMEC. In replicate experiments with HMEC from three individuals, we determined that 2 Gy X-irradiation causes increases in the appearance of p16(-) cells with long term growth potential. Flow cytometry confirmed that the differences in BrdU incorporation in the treated and untreated populations were statistically significant (P = 0.0016). Based on these data, we hypothesize that radiation to cause epigenetic changes in p16 expression. We are currently exploring the mechanistic underpinnings of this phenomenon.

We have demonstrated that irradiated pre-malignant HMEC, cultured in reconstituted basement membrane and transforming growth factor beta1 (TGFbeta), fail to undergo tissue-specific morphogenesis when embedded in reconstituted basement membrane, which is strongly associated with malignancy and deregulation of differentiation. Our recent studies show that disrupted morphogenesis is a result of irradiated cells undergoing TGFbeta mediated epithelial to mesenchymal transition (EMT). These data are now in press in Cancer Research. We compared eqi-toxic doses of 1Gy of 1 GeV/amu 56Fe to 2Gy of 137Cs gamma-radiation in both monolayer and 3D culture. Both exposures resulted in EMT in monolayer and disrupted morphogenesis when cells were cultivated in the presence of TGFbeta. Radiation dose responses show that this effect has a very low threshold in that a single exposure of 3-100 cGy gamm-radiation elicits the ‘same’ phenotypic switch.

Fe ions were more toxic than x-rays to HMEC irradiated in growth arrested monolayer cultures, with a RBE of 1.8 at the D(10) level with a delayed plating protocol. Irradiation as growth arrested acinar structures protected HMEC against x-ray-induced cell death. In contrast, the 3D architecture did not protect the HMEC from killing by Fe ions. Exposure of HMEC in growth arrested monolayer culture to x-rays or Fe ions elicited severe karyotypic instability, defined as at least 3 subpopulations of cells per colony, in the clonal descendants of ~20% of clones. This phenotype was not seen in any of the unirradiated colonies. Severe karyotypic instability was seen in colonies arising after exposure to the lowest dose of Fe ions (0.5 Gy), but no evidence of severe karyotypic instability was seen for colonies arising after exposure to <2 Gy of x-rays. There was no evidence of a dose response for severe karyotypic instability. These data are now submitted.

A dose response was seen for centrosome aberrations (CA) defined as 3 or more centrosomes/cell for the progeny of cells exposed in monolayers to x-rays or Fe ions. Unirradiated clones had a baseline frequency of 0.92% CA. The dose response appeared to saturate with ~4% of cells/colony exhibiting this phenotype. There was no evidence for an RBE>1 for the induction of centrosome hyperamplification following irradiation in 2D. However, irradiation in rBM reduced the level of centrosome abnormalities after x-irradiation, but did not protect against Fe ion exposure, suggesting that there may be an increased RBE for centrosome abnormalities in cells exposed to Fe ions.

Studies have analyzed gene expression induced by low-LET radiation in growth arrested HMEC using statistical and bioinformatics approaches with the goal of identifying specific signaling pathways. Image analysis continues to develop strategies that can be used to characterize the diversity of cell responses and track the development of phenotypes over time.

Together these studies show that in general high vs low LET radiation is more effective in eliciting many of these endpoints. We propose that these studies warrant further examination of whether individual phenomena (e.g. damage response, phenotype, genomic instability) interact in a unique manner to drive neoplastic transformation following high LET exposure. We have begun a new project in collaboration with the Boston NSCOR to model these events at the cellular level.

IEEE International Symposium on Biomedical Imaging: from Nano to Macro, 2007, pp. 532-536. ISBN: 1-4244-0672-2 http://dx.doi.org/10.1109/ISBI.2007.356906 , Apr-2007

IEEE International Symposium on Biomedical Imaging: from Nano to Macro, 2007. pp. 524-527. ISBN: 1-4244-0672-2 http://dx.doi.org/10.1109/ISBI.2007.356904 , Apr-2007

Berlin : Springer, 2006. pp. 586-595. (Advances in Visual Computing ; vol. 4292) http://dx.doi.org/10.1007/11919629_59 , Nov-2006

To provide a comprehensive picture of HZE effects from the initial damage, to early cellular HMEC responses, to persistent functional precursors of carcinogenesis.

Several NASA areas of interest will be addressed: mechanisms, nature, and frequency of DNA damaging events; mechanisms of DNA repair and misrepair; early signal transduction mechanisms; immediate and long-term, and reversible and irreversible gene expression changes; cellular remodeling and reorganization; potential mechanisms of tissue “repair” and matrix effects; induction and regulation of genomic instability; and the cellular and molecular mechanisms of charged particle-induced progression to a neoplastic phenotype. The data obtained in these studies will be integrated at two levels: by theoretical modeling of the physical events leading to DNA damage and by systems biology modeling of critical pathways. Ultimately, this work will help NASA develop a basic understanding of the mechanisms of HZE radiation damage and repair and their contributions to the neoplastic process.

In the context of the NSCOR, we have developed image analysis programs to quantify DNA damage response (e.g. radiation-induced foci) and specific organelles (i.e. centrosomes) using high-throughput imaging bioinformatics.

Mechanisms of HZE Damage and Repair in Human Epithelial Cells

PI: Barcellos-Hoff

Co-PI: Chatterjee, Costes, Kronenberg, Parvin, Rydberg, Yaswen

Specific Aim 1. Determine whether rapid phenotypic responses of HMEC induced by HZE are more complex and prolonged than those induced by X-rays.

Rydberg and Groesser : We have performed studies to examine the persistence of gamma-H2AX, 53BP1 and ATM 1981p foci as indication for persistent DSBs in HMEC cells. Experiments with 184v HMEC in monolayer cultures were carried out with 1 GeV/n Fe ions (LET 150 keV/µm) at doses of 0.5 Gy and 1 Gy. Such doses are small enough to allow the cells to proliferate with minimal cell cycle delays. This contrasts with most published work addressing persistent DSBs, which were carried out with doses of 40 - 80 Gy resulting in permanent G2 block.

We find that 24 hr after exposure, many more gamma-H2AX foci remain after Fe irradiation than after gamma irradiation. The result is consistent with the hypothesis that Fe ion exposure causes more resistant, difficult-to-repair DSBs than low LET gamma radiation. Although this is in agreement with previous high-dose experiments in other cell lines, it was important to verify this at low doses in our proliferating cells. However, it remains an open question whether all the foci at late time-points after irradiation necessarily reflects DSBs. In contrast to ?-H2AX foci, we find a return of 53BP1 foci to background level at late time points. This discrepancy may either indicate that the gamma-H2AX foci do not represent DSBs or it could indicate that complex difficult-to-repair DSBs are nor recognized by 53BP1. Further experiments are needed to investigate this question.

Costes and Barcellos-Hoff: As noted last year, we recently published observations in human fibroblasts exposed to low-LET radiation that show a clear discrepancy between RIF counts and expected number of DSB (1). To better define the relationship between physical damage and biological context in radiation-induced foci (RIF) following HZE radiation exposure, we used a well-accepted physical model of the track structure for HZE particles developed by Ponomarev and Cucinotta to simulate the production of DSB in a modeled nucleus in comparison to experimental RIF induced by HZE. In this approach, the three-dimensional space was divided into cubical pixels of the size corresponding to that of microscope image pixels. DNA was arranged into two types of intermittent bands: dense regions of DNA based on a random-walk geometry (heterochromatin), and low-density homogenous regions (euchromatin)in the simulated nucleus. The model images were then blurred to create more realistic images by applying a Gaussian filter, which takes into account the point-spread function of the microscope. We then generated nuclear images with the frequency of DSB per pixel. Applying the Gaussian blurring to these images produced images with foci-like objects, which we refer to as pseudo-foci (pRIF). The simulations yielded 1.25 DSB/µm and 0.8 pRIF/µm along 1 GeV/amu Fe track. As expected from the theoretical predictions the simulation generated pRIF preferentially in the denser regions of DNA and randomly located along ion tracks.

Monte Carlo simulation predicted that the initial distribution of pRIF along HZE tracks would be random. In contrast, our analysis of RIF data collected at NSRL-6 and 7 observations showed some regularity in RIF spatial distribution even at the earliest time point. We also determined that RIF were more frequently located in euchromatin than predicted by modeling and occurred predominantly at the interface of euchromatin and heterochromatin for both low and high LET. We thus hypothesize that the type of chromatin organization where DSB are generated dictates the kinetic of RIF formation and may influence DNA repair.

Specific Aim 2: Determine whether persistent phenotypic responses of HMEC induced by HZE are more complex and prolonged than those induced by X-rays.

Barcellos-Hoff and Andawarra: The NSCOR invited Dr. Les Redpath of UC Irvine to spend a day discussing the applications and limitations of current human models of carcinogenic transformation by ionizing radiation. We have used the ability of human mammary epithelial cells (HMEC) to undergo acinar morphogenesis in a reconstituted basement membrane to show that radiation exposure of individual cells leads to a persistently altered phenotype in daughter cells that alter epithelial interactions and the microenvironment (2). An important implication from these studies is that the irradiated epithelial phenotype lacks critical controls imposed by the microenvironment to maintain tissue, and potentially genomic integrity. We have generated pilot data using embedded cultures for which we analyze morphology, acinar size, expression of mesenchymal markers and epithelial markers such as E-cadherin and ß-catenin using immunofluorescence and protein blot analysis. The preliminary analysis of these specimens using confocal or deconvolution microscropy indicate that HZE is as effective as sparsely ionizing radiation in eliciting the irradiated phenotype of disrupted morhogenesis. We have now determined that irradiated HMEC exposed to TGF-ß are persistently dysfunctional in a manner consistent with epithelial to mesenchymal transition (EMT). We performed studies at the NSRL-7 using 184 HMEC, which are finite lifespan post-selection HMEC from one specimen, 184. We repeated the monolayer culture experiments that were performed at NSRL-5 comparing the phenotype of cells exposed to equitoxic doses of 1 GeV Fe or gamma-rays. HMEC underwent comparable degrees of EMT in response to either densely and sparsely ionizing radiation at equitoxic doses. Additional studies will be done at LBNL to understand the molecular mechanisms that promote TGF-ß induced EMT in HMEC. Yaswen and Mukhopadhyay: As part of our effort to develop assays to distinguish differences in the persistent biological effects of high and low LET radiation, we examined the ability of HMEC to escape the first barrier to immortality, which is called stasis. Many laboratories have found that it is exceedingly rare for normal human epithelial cells to escape the mortality barrier, called agonescence (3), to generate immortal cell lines. Thus, in these experiments, fourth passage (4p) pre-stasis HMEC derived from primary cultures grown in serum-free MEGM medium were exposed to a single 2 Gy dose of X-rays. Replicate cultures of the irradiated and sham-irradiated cells in three independent experiments using HMEC from two individuals, developed patches of small actively proliferating cells in the cultures derived from irradiated cells only. Ultimately, after considerable delay, a few patches of small actively proliferating cells appeared in the control dishes as well. However, the irradiated cultures continued to proliferate after 40 additional population doublings, while the sham-irradiated cells ceased net proliferation after 15 additional population doublings. Staining for p16 protein and SA- gal activity confirmed that 2 Gy irradiation caused striking increases in the frequency and rate of appearance of p16(-) SA- gal(-) cells with long-term growth potential. Based on these data, we hypothesize that radiation, possibly acting through the generation of reactive oxygen species, causes this epigenetic change. Using this new assay, quantification of the relative biological effectiveness of high LET particles (1 GeV Fe - 1 Gy) compared to low LET reference radiation is currently being established using pre-stasis HMEC irradiated at BNL during NSRL-8.

Aim 3 Assess whether a persistent state of genomic instability is induced in the progeny of HMEC exposed to HZE or X-irradiation.

Kronenberg, Sudo, Barcellos-Hoff, Gray: This year the focus has been on the assessment of genomic instability in finite lifespan 184v HMEC using multi-color whole chromosome FISH painting to address the incidence of non-clonal structural chromosome rearrangements, and centrosome staining to monitor the fraction of cells within a colony that display supernumerary centrosomes. All experiments have been performed with colonies that survive exposure to graded doses of x-rays or 1 GeV/amu Fe ions from cells irradiated as monolayers.

G-banding analysis of un-irradiated 184v cells demonstrates a stable karyotype of 46 XX, with no deviation from this chromosome complement. Chromosome painting analysis of the progeny of un-irradiated single cells also demonstrates the stability of the karyotype, with no evidence of non-clonal structural rearrangements. In contrast, both x-ray exposure (1-6 Gy) and Fe ion exposure (0.5-3 Gy) result in a substantial fraction of clones with two or more karyotypes, indicating that there is a memory of the irradiation exposure that is transmitted to the surviving daughter cells and that is expressed as a variation in the structural assortment of the chromosomes. To our knowledge, this is the first demonstration of chromosomal instability in finite-lifespan human mammary epithelial cells exposed to ionizing radiation. The transition to genomic instability has been associated with the progression from usual ductal hyperplasia to ductal carcinoma in situ in human biopsy materials (4).

Increased frequency of supernumerary centrosomes may increase the probability of abnormal segregation of chromosomes due to multipolar mitotic spindles. This latter phenotype has been found in human mammary biopsy materials with ductal carcinoma in situ, and was associated with the development of chromosomal instability (5). There is a low but detectable level of centrosomal instability in the clonal descendents of un-irradiated HMEC clones (~1% of cells in a discrete colony have 3 or more centrosomes). Assessment of radiation-induced centrosome abnormalities in 184v cells exposed to either x-rays or Fe ions indicates a dose-dependent increase in supernumerary centrosomes in the clonal progeny of irradiated cells for each radiation type. The results for both x-rays and Fe ions are similar as a function of dose. The vast majority of clones from x-irradiated cells demonstrate an increase in supernumerary centrosomes that is > 3 S.E. above the level found in the un-irradiated clones. The clonal progeny of Fe-irradiated cells also demonstrated a similar increase in the fraction of clones with increased levels of supernumerary centrosomes.

The experiments to date indicate that exposure to either x-rays or Fe ions results in the transmission of a memory of that exposure that can be expressed many cell divisions later in the descendants of single irradiated cells. Intraclonal karyotypic heterogeneity may increase the probability of development of variants that can advance the neoplastic phenotype. Thus, exposure to both sparsely and densely ionizing radiation can result in genomic instability in the progeny of irradiated human mammary epithelial cells with finite lifespan, and elicits a series of responses that have been associated with the transition to neoplasia in human biopsy material.

Specific Aim 4: Define theoretical and system biology models of HZE damage and response relevant to carcinogenesis.

Costes, Chatterjee and Boissiere: A focus of this project is to develop theoretical models for carcinogenesis. The initial approach was based on the assumption that the mutation rate represents an "initiating event" followed by clonal expansion rate for cells that carry mutation, and subsequently acquire several other mutations either directly from irradiation, spontaneous processes or through induced instability or mutator phenotype caused by an earlier mutation, etc. This approach relates to parametric fitting of data and may not be the ultimate choice as a model for carcinogenesis. The tendency in radiation biology is to model events as hierarchical, linear responses, rather than combinatorial networks. Thinking of radiation responses as assets of a network, rather than programmed responses of cells, could be used to understand the dynamic flexibility of the network. New modeling approaches will be required to represent the non-linear aspect and adaptive behavior of system biology.

To this end, we convened an informal meeting on modeling was held in Berkeley between the LBNL NSCOR and the Tufts NSCOR led by Dr. L. Hlatky, as well as Dr. R.J.M. Fry and Dr. F. Cucinnotta, to discuss how one would incorporate the effect of the microenvironment into carcinogenic risk model. The group discussed current experimental models basis and the potential for an integrated model using well-quantitated radiation responses using systems biology principles of network interconnectivity and spatial organization of cellular phenotypes within the higher order multicellular structure (6). We believe that by improving radiation carcinogenesis models in a progressive manner by incorporating more biological mechanisms in a more realistic way will help inform mechanism based modeling of risk.

Parvin and Das: In support of systems biology, algorithms have been developed for quantifying biological properties of imaging assays for both monolayer as well as embedded cell culture models. The emphasis has been on robust geometric models and improved accuracy of capturing pertinent biological properties. These computed representations are registered with a database that maps them to experimental variables and their raw representation. Additionally, a number of invariance and information preserving features are routinely computed, which can aid in constructing predictive models.

In a parallel effort, we have developed algorithms to identify the cis-regulatory elements that direct gene expression by integrating nucleotide sequence and microarray expression data sets. We have used adaptive modeling strategies to account for complex dependencies at both transcriptional and post-transcriptional levels. The central focus has been on adaptive regression splines, which are also closely related to switch-like behavior of the response function. These algorithms are currently being applied to decipher the transcriptional endpoints of TGF-ß mediated signaling pathways in irradiated cells and the elements of transcriptional regulatory logic that discriminate between monolayer or embedded culture.

REFERENCES

1. Costes, S.V., Boissiere, A., Ravani, S.A., Romano, R., Parvin, B., and Barcellos-Hoff, M.H. 2006. Imaging features that discriminate between high and low LET radiation-induced foci in human fibroblasts. Radiat Res 165:505-515.

2. Park, C.C., Henshall-Powell, R., Erickson, A.C., Talhouk, R., Parvin, B., Bissell, M.J., and Barcellos-Hoff, M.H. 2003. Ionizing Radiation Induces Heritable Disruption of Epithelial Cell-Microenvironment Interactions. Proc Natl Acad Sci 100:10728-10733.

3. Romanov, S.R., Kozakiewicz, B.K., Holst, C.R., Stampfer, M.R., Haupt, L.M., and Tlsty, T.D. 2001. Normal human mammary epithelilal cells spontaneously escape senescence and acquire genomic changes. Nature 409:633-637.

4. Chin, K., de Solorzano, C.O., Knowles, D., Jones, A., Chou, W., Rodriguez, E.G., Kuo, W.L., Ljung, B.M., Chew, K., Myambo, K., et al. 2004. In situ analyses of genome instability in breast cancer. Nat Genet 36:984-988.

5. Lingle, W.L., Barrett, S.L., Negron, V.C., D'Assoro, A.B., Boeneman, K., Liu, W., Whitehead, C.M., Reynolds, C., and Salisbury, J.L. 2002. Centrosome amplification drives chromosomal instability in breast tumor development. PNAS 99:1978-1983.

6. Barcellos-Hoff, M.H., and Costes, S.V. 2006. A systems biology approach to multicellular and multi-generational radiation responses. Mutation Res 597:32-38.

Advances in Visual Computing, First International Symposium, ISVC 2005, Lake Tahoe, NV, USA, December 5-7, 2005, Proceedings. p. 427-436, 2005. , Dec-2005

Costes and Barcellos-Hoff We established an imaging algorithm that identifies radiation-induced foci (RIF) automatically in the nucleus after exposure to radiation in collaboration with Dr. Parvin. This algorithm was developed and optimized on human diploid fibroblast grown in deep confluence. Our analysis is under review for publication at Radiation Research and was presented as a poster at the Gordon Conference in Radiation Oncology in February 2005, the NASA meeting in May 2005, and the Radiation Research Society in October, 2005. We are now using similar algorithms to study 184 HMEC after exposure to 1 GeV/amu Fe and X-rays. For that purpose, we irradiated HMEC184 during NSRL4 with a 1 Gy dose, with the beam parallel to the dish where the cells were grown. With such geometry, Fe particle traversal could be visualized. Cells were fixed at 10, 30, 60, 120, 180 and 300 min post irradiation and immunostaining was performed for different DNA damage markers (i.e. ?H2AX, ATMp and 53BP1). These data are currently analyzed to study the spatial distribution of RIF in the nucleus as well as the spatial correlation of RIF with DNA density. This study has been recently extended to HMEC184 grown in three dimensions. For that purpose a new imaging method based on search of local maxima was developed to determine the position of RIF. Figure 1 illustrates the accuracy of this method in an acinus irradiated by 2 Gy of X-rays. Spatial localization of RIF in the nucleus will be studied using this tool in a similar manner it is currently done in two-dimension. These results will be presented as a poster at the next Radiation Research Society meeting in October 2005. Finally, HMEC184 grown in 3D or 2D will be irradiated by 1 GeV/amu Fe at the next BNL run (October-November 2005) and images will be acquired and analyzed in the next two to three months following the run.

Specific Aim 2: Determine whether persistent phenotypic responses of HMEC induced by HZE are more complex and prolonged than those induced by X-rays. Barcellos-Hoff and Andawarra Studies will be performed at the NSRL-7 using 184 HMEC, which are finite lifespan post-selected HMEC from one individual, specimen 184. These cells retain most normal HMEC biology, but have overcome a stress-associated senescence barrier (stasis) via silencing of the cyclin-dependent kinase inhibitor p16INK4a. We will repeat the same 2D experiments that were performed at NSRL-5, in order to reconfirm our data. Other than that we will be carrying out some pilot experiment in 3D with Fe (1 Gy) or ?-rays (2Gy). And we will be analyzing for morphology, expression of mesenchymal markers such as N-cadherin, fibronectin and vimentin, and epithelial markers such as E-cadherin, b-catenin and ZO-1, using immunofluorescence and protein blot analysis. Additional studies will be done at LBNL to understand the molecular mechanism of TGF-ß induce epithelial to mesenchymal transition (EMT) in irradiated HMEC. Some of our preliminary data have already been /will be presented at the AACR-pathobiology workshop, July17-25, 2005, Aspen, Colorado and at Radiation Society Meeting, October 15-19, 2005, Denver, Colorado.

Yaswen and Mukhopadhyay Post-selection 184 HMEC were used to evaluate the potential of X-rays to cause reactivation of telomerase and immortalization – potentially rate-limiting steps in the early development of human breast cancer. These experiments were conducted using treatment regimens in which both the dose and dose rate were varied. We found that post-selection HMEC did not show evidence of X-ray-induced changes in telomerase expression and immortalization. No immortal clones were generated from a total of 3.3 x 106 cells irradiated. This negative result is consistent with that of many other studies that have shown human cells are extraordinarily resistant to oncogenic transformation in culture. Thus the incidence of X-ray-induced immortalization of untreated post-selection HMEC will likely be too low to measure accurately to provide a benchmark for future studies of HZE. Having anticipated this possible result, we are proceeding with our plan to use genetically modified HMEC in which one or more predisposing defects are engineered into targeted HMEC. As stated in our original proposal, the rationale for using such genetically modified HMEC is that the successful detection of an agent that contributes to elongated or indefinite life span depends on the use of a target cell population where there is a good likelihood that only a single change is necessary to alter replicative potential. Since we hypothesize that multiple cooperative changes are necessary for post-selection cells to overcome telomere dysfunction-associated agonescence and progress to immortality, we are providing some of the known contributing changes in the form of retrovirally transduced oncogenes such as c-MYC and/or ZNF217.

We have performed one experiment in which we irradiated 30.4 x 106 HMEC constitutively over-expressing ZNF217 with 2 Gy X-rays, and have concluded that even in this case, immortalization is a rare event. This data indicates that other cooperative changes (e.g., increased c-MYC expression and/or p53 inactivation) are required. Since some immortal HMEC lines generated after transduction with both ZNF217 and c-MYC show no evidence of regional DNA-sequence copy number variations by array comparative genomic hybridization (unpublished data), it is possible that the levels of ZNF217 and/or c-MYC expression must reach certain levels or be within certain limits for immortalization to occur. Alternatively, additional epigenetic changes or small genetic events may be required. We are currently determining whether X-rays will increase the frequency of immortalization of HMEC transduced with both ZNF217 and c-MYC. A genetic suppressor element (GSE) that effectively abrogates p53 function is also being used in these studies. In addition, we will determine whether the presence of hTERT, the catalytic subunit of telomerase, can be used as a sensitive surrogate marker of immortalization. This marker may be assayed more readily than immortalization itself, and may facilitate more quantitative measurements of immortalization frequency.

We are also determining the effects of X-irradiation and HZE on anchorage-independent growth (AIG) of immortalized HMEC, as this easily quantifiable phenotypic trait has long been associated with malignant transformation. Alterations in the PI3K and TGF-? signaling pathways are known to influence AIG. The susceptibility of immortalized HMEC with different genetic backgrounds (e.g., p53+ vs. p53-) and in different microenvironments (e.g., in the presence or absence of TGF-?) to radiation-induced AIG is being assessed with the objective of developing a sensitive assay of persistent radiation effects with direct relevance to malignant transformation.

Aim 3 Assess whether a persistent state of genomic instability is induced in the progeny of HMEC exposed to HZE or X-irradiation. Kronenberg, Sudo, Gray Cells in G1 phase possess a single centrosome that replicates to help form the mitotic apparatus and ensure normal segregation of chromosomes. Cells with abnormal numbers of centrosomes (i.e. 3 or more) are at elevated risk for the development of aneuploidy, and centrosomal instability has been noted as a hallmark of cancer, including breast cancer (5). Although there are many reports of radiation-induced genomic instability, the preponderance of the data in the literature involves immortalized cells (reviewed in (6)). The reports in human cells with a finite lifespan are limited primarily to fibroblasts and lymphocytes (reviewed in (7)). Progress on the incidence of genomic instability in the progeny of irradiated HMEC has focused on the collection of individual surviving colonies of finite lifespan HMEC in monolayer culture that were treated either with x-rays or 1 GeV/amu Fe ions. HMEC 184v cells were used at early passages (passage 6 or passage 7) for these experiments. Cells were irradiated in G0 phase and dispersed for colony forming assays Eighty four clones were isolated from cells seeded after exposure to x-rays, with an additional 27 clones isolated from the parallel control cultures. Thirty four of these clones have been processed to determine the fraction of cells within a clone that exhibit centrosome abnormalities. The remaining 77 clones have been prepared for fluorescence in situ hybridization analysis of intraclonal karyotypic heterogeneity, another indicator of genomic instability. During NSRL-5 and NSRL-6 119 clones were collected that survived irradiation with densely ionizing 1 GeV/amu Fe ions, and an additional 29 clones from parallel control cultures. Of these, 73 clones have been designated for the analysis of centrosome abnormalities, and 98 for the analysis of intraclonal karyotypic heterogeneity.

Finite-lifespan HMEC 184v cells exposed to x-rays in monolayer culture showed a dose-dependent increase in the fraction of cells within a clone that have 3 or more centrosomes/cell. A small fraction of cells from un-irradiated clones (<1.0%) had abnormal numbers of centrosomes. The dose-response for the induction of centrosomal abnormalities in the x-irradiated clones analyzed to date is described by a linear quadratic equation (r2=0.989). Clones collected during NSRL-5 and NSRL-6 are being evaluated to assess the effectiveness of densely ionizing Fe ions for the induction of centrosomal abnormalities. Studies are also underway on x-irradiated and Fe ion exposed clones to evaluate intraclonal karyotypic stability using chromosome painting approaches.

Specific Aim 4: Define theoretical and system biology models of HZE damage and response relevant to carcinogenesis.

Chatterjee and Boissiere We have been directing our efforts in two areas of research: biological significance of DNA damage clusters and considerations of various biological processes related to carcinogenesis prior to developing a reliable theoretical model. (i) Biological Significance of DNA Damage Clusters: We have now extended our previous studies of correlating clusters of damage with local hprt mutations. In these studies we have made a very simple assumption that the theoretical mutation frequencies are exactly equal to the cluster frequencies based on the model calculations. We have found that for 50 MeV protons, clusters with one or more double strand breaks and two or more associated base damages have equivalent mutation frequencies that match with the experimental value of 1.72E-08. The criterion for cluster characteristic remains the same for 1000 MeV/n iron particles. In addition, it seems that under the assumption that the cluster frequency and mutation frequency are equivalent, the cluster classification with 5 or more base damages without any double strand break, also yield results which are in agreement with experimental data for both qualities of radiation. (ii) Approaches to developing a Theoretical Model for Carcinogenesis: Our initial approach depends upon the assumption that the mutation rate represents an “initiating event” followed by clonal expansion rate for cells that carry mutation, and subsequently acquire several other mutations either directly from irradiation, spontaneous processes or through induced instability or mutator phenotype caused by an earlier mutation, etc. This approach relates to parametric fitting of data and may not be the ultimate choice as a model for carcinogenesis. However, we believe that we will learn how to improve the model in a progressive manner by incorporating more biological mechanisms relevant to carcinogenesis and perhaps in a more realistic way.

PRESENTATIONS 1.Invited Talk: Simulated Space Radiation Studies for The Assessment of Chromosomal Damage: An Integrated Experimental and Theoretical Research at the 16th NASA Annual Meeting, April 2005. 2. Poster presentation titled “Elucidating Radiation Effects on Centrosomal and Chromosomal Stability in Human Mammary Epithelial Cells“ at the MICRO2005, November13-18, 2005, Venezia, Italy. 3. Invited Talk: Development of Theoretical Models in Radiation Biology With Special Emphasis on HZE Particles at the MICRO2005, November 13-18, 2005, Venezia, Italy. Parvin and Das Two different facets of Systems Biology are being explored: (1) statistical correlation between heterogeneous datasets and (2) systematic understanding at the transcription level. The focus of statistical correlation remains on quantitative assessment of each data modalities (e.g., immunofluorescence, gene expression) followed by correlation across multiple datasets. Toward this objective, a system has been developed for quantitative assessment of DNA damage and centrosome abnormalities. This system is referred to as the BioQuant_2D_Assay. In each case, the nuclear morphology provides the context for detailed protein localization studies. The previous system, reported last year, has been extended to localize nuclear regions based on convexity, partitioning overlapping nuclear compartments based on grouping of self-similar regions from curvature features, and quantification of protein co-localization events through geometric analysis. Nuclear regions provide the context for co-localization of specific proteins. The main innovation is development of model-based techniques for delineation of nuclear regions as well protein localization. The BioSig imaging bioinformatics system has been redesigned and being evaluated for streamlining data acquisition, processing, annotation, and visualization. The main innovation is extensive features facilitating usability through standardized controlled vocabularies that are derived from MAGE ontologies and coupled with Open Microscopy Environment (OME). The new architecture facilitates data mining and protein co-localization across many studies, where each study is composed of multiple experiments. With respect to the mechanistic understanding of radiation responses, we have initiated a study into how phenotypic complexity is regulated at the transcription level. Regulation of gene transcription in eukaryotes is complex and is inherently combinatorial in nature. Cooperative binding of transcription factors to the promoter DNA is a key element of such combinatorial control in gene regulation networks. Current computational models do not take adequate account of this synergistic activation and repression. We had previously built a tool based on Multivariate Adaptive Regression Splines (MARS) that overcomes these limitations. MARS builds the response function in terms of piecewise linear functions and their products. Thus, interactions and non-linearities in the data are captured in the simplest non-parametric fashion. Application of this method to yeast cell-cycle microarray expression data has showed that it can achieve accuracy 1.5-3.5 times that of the current computational methods, while at the same time can predict known motifs and their cooperative pairs at appropriate stages of the cell-cycle. We plan to apply the mammalian version of this technique to discover the transcriptional sub-networks that are activated/deactivated in response to different radiation exposures or Tgf?1 gene status. By transcriptional sub-network, we mean active cis-regulatory motif combination, its direct targets and the physiological processes it regulates. We will scan a large set of candidate position weight matrices to obtain a prioritized set and then run MARS on this prioritized set to infer the functional elements and their cooperative combinations. Previous application of this method to adult human liver microarray data showed that this approach can uncover known transcriptional sub-networks. In this application it also uncovered several novel sub-nets, which share similar functional characteristics as the known ones. Additionally, regulatory cross-talk was also detected among multiple tissues, along the notion of synexpression groups. We expect that this technique will lead to a comprehensive understanding of transcriptional networks in irradiated mammalian cells. With respect to quantitative characterization of immunofluorescence, the following advancements are planned: (1) Routine utilization of BioQuant_2D_Assay analytical system for quantitative characterization of centrosome abnormality and foci formation, (2) development of BioQuant_MultiCellular_ECM_Assay for characterizing cell-cell communication and cell-ECM interaction, (3) development of BioQuant_MultiCellular_3D_Assay for foci characterization, and (4) routine utilization of BioSig Imaging Bioinformatics system for uploading images and viewing experimental data. While quantitative assessment of cell-cell and cell-ECM interaction is will be characterized from 2D images, foci formation will be characterized from three-dimensional datasets.

With regard to the systematic understanding of regulatory mechanisms, following advancements are planned: (1) deciphering the regulatory subnetworks activated by TGF-beta. To this end, we are studying the change in regulatory architecture when TGF-beta is knocked out or overexpressed. Initial results are indeed supported by available evidence. We expect to finalize our study in the coming months. (2) how the pathways activated by TGF-beta are affected upon irradiation; (3) how aberrant regulation of splicing leads to malignant phenotype; and (4) development of more effective quantitative methods to decipher regulatory controls.

Apr-2005

Nov-2005

Oct-2005

Oct-2006

Oct-2005

Jan-2005

Nov-2005

May-2005

Participants in this NSCOR have developed and implemented unique cell culture methodologies that more accurately reflect the behavior of human epithelial cells in vivo. In the first year of the program, we have recruited 4 new staff members who have been trained to execute and analyze HZE radiation response using the HMEC models at NSRL 3 and 4. Cell culture and gene expression core resources were established. The data from the first two NSRL runs have determined the relative biological effectiveness for 90% clonal death. In addition, we have developed techniques and tools for imaging nuclear foci localized by immunofluorescence as a measure of DNA damage and repair. Theoretical modeling of DNA damage indicate that there are many more complex DNA aberrations from HZE radiation exposure than what is measured by experiments. These results suggest that underestimation of this damage may be a significant factor in risk estimations.

The objective for the coming year is to use system biology to build an integrative model of data on the rapid and persistent responses to HZE radiation damage.

May-2004

Feb-2004

Jan-2004

Feb-2004