LET Dependence of Radiation-Induced Bystander
Effects using Human Prostate Tumor Cells
by
Vered Anzenberg
M.S., Nuclear Science & Engineering
Massachusetts Institute of Technology, Cambridge, MA 2005
B.S., BioNuclear Engineering
University of California Berkeley, Berkeley, CA 2003
SUBMITTED TO THE DEPARTMENT OF NUCLEAR SCIENCE & ENGINEERING IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE OF
DOCTOR OF PHILOSOPHY OF NUCLEAR SCIENCE & ENGINEERING
AT THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
June 2008
© Massachusetts Institute of Technology. All rights reserved.
The author hereby grants to MIT permission to reproduce
and distribute publicly paper and electronic
copies of this thesis document in whole or in part.
Author .............................
.......................................................
Department of Nuclear Science & Engineerinp
May 15, 2008
Certified by..............................
SJeffrey
A. Coderre
Associate Professor of Nuclear Science & Engineering, MIT
Thesis Supervisor
C ertified by ..................................
.........................
(It/
'/ cquelyn C. Yanch
Professor of Nuclar Science & Engineering, MIT
Thesis Reader
Read by.........................................
.............................
Bevin P. Engelward
Professor of Biological Engineeri'g, MIT
r
~/v /f The~e Reader
A ccepted by ............................................................. ...
Chairman
DepartmentalquelynC.Yanch
MA
HUSETTS NSTT
OF TECHNOLOGy
JUL 2 2008
LIBRARIES
Chairman, Departmental Comttee on Graduate Students
LET Dependence of Radiation-Induced Bystander
Effects using Human Prostate Tumor Cells
by
Vered Anzenberg
Submitted to the Department of Nuclear Engineering on May 15, 2008 in partial
fulfillment of the requirements for the Degree of Doctor of Philosophy in
Nuclear Science & Engineering
Abstract
In the past fifteen years, evidence provided by many independent research groups
have indicated higher numbers of cells exhibiting damage than expected based on the
number of cells traversed by the radiation. This phenomenon has been coined as the
"bystander effect". The purpose of this study was to characterize the ability of irradiated
tumor cells to induce bystander effects in co-cultured cells. Human DU-145 prostate
carcinoma cells were grown on a 1.4 pm-thick mylar membrane in specially constructed
cell culture dishes for irradiation with alpha particles (average energy 3.14 MeV) from a
241Am source, or in 6-well plates for irradiation with 250 kVp x-rays at 25oC. In parallel
experiments, the tumor cells were incubated at 40 C for one hour prior to irradiation and
irradiated on ice to test the nature of the bystander signal. Bystander cells were placed
into the medium above the irradiated DU-145 and were co-incubated for a length of time.
The bystander effect endpoints measured in either DU-145 tumor cells or in normal
primary AGO1522 fibroblasts were micronucleus (MN) formation, y-H2AX double strand
break repair foci, and survival fraction. A 1.5-2.0-fold increase in MN formation was
observed in both DU-145 and AG01522 bystander cells after either alpha particle or xray irradiation of the DU-145 target cells. A 1.5-fold y-H2AX bystander increase and a
survival fraction reduction to 80% were only detected in AGO1522 cells, and only after xray irradiation of target DU-145 cells. Alpha particle irradiation of the target DU-145
cells produced neither y-H2AX foci nor survival fraction bystander effect in either cell
line. Lowering the temperature to 4oC during the irradiation of the DU-145 tumor cells,
with either x-rays or alpha particles, eliminated both the MN formation and the decreased
survival fraction bystander effects in the co-cultured AG01522 fibroblasts. This study
demonstrates that biochemical processes in the directly-irradiated tumor cells are required
for initiation of the signaling process. Low temperature during the irradiation inhibited
the initiation of a bystander signal. There are also LET-dependent differences in the
signal released from DU-145 human prostate carcinoma cells; and that, for some
endpoints, bystander AG01522 fibroblasts and bystander DU-145 prostate carcinoma
cells respond differently to the same, medium-mediated signal.
Thesis Supervisor: Jeffrey A. Coderre, Ph.D.
Associate Professor of Nuclear Science & Engineering, Massachusetts Institute of Technology
Acknowledgments
Cloths of Heaven
Had I the heavens' embroidered cloths,
Enwrought with golden and silver light,
The blue and the dim and the dark cloths
Of night and light and the half-light,
I would spread the cloths under your feet:
But I, being poor, have only my dreams;
I have spread my dreams under your feet;
Tread softly because you tread on my dreams.
- William Butler Yeats
I would like to first thank my advisor, Professor Jeffrey A. Coderre for his constant
guidance and encouragement through out my years at MIT. I would like to thank
Professor Jacquelyn C. Yanch for being an amazing mentor. I would like to especially
thank Dr. Hongying Yang, who taught me all the biological experimental techniques that
I know but most importantly, she taught me to believe in myself, and my abilities as a
scientist. I would like to thank Dr. Martin Purschke for his generous help with the cell
cycle analysis.
To my family, you are everything. To my friends, you know who you are, and you know
how much I value our friendship and our time together.
This research was performed under appointment to the U.S. Department of Energy
Nuclear Engineering and Health Physics Fellowship Program sponsored by the U.S.
Department of Energy's Office of Nuclear Energy, Science, and Technology.
Contents
1 Introduction
12
1.1 DNA Damage in Directly Irradiated Cells.................
.........................17
1.1.1
DN A Repair.......................
1.1.2
Cell Irradiation at Low Temperatures.....................................21
..............................................19
1.2 The Bystander Effect......................................................................... 23
1.2.1
In vitro Bystander Effect Studies....
................
............................
24
1.2.1.1 Gap-junction Mediated Bystander Effects..............................25
1.2.1.2 Medium-Mediated Bystander Effects .........
................... .. 27
1.2.1.3 Microbeam Bystander Effects Studies ....................................... 30
1.2.1.4 DNA Repair in Bystander Cells...........................................35
1.2.1.5 Apoptosis in Bystander Cells....................................
....... 42
1.2.1.6 The Adaptive Response and Bystander Effects .......................... 48
1.2.1.7 LET Dependence of the Bystander Effect....................
....... 50
1.2.1.8 Bystander Effects in Tumor Cells............................................56
1.2.1.9 Temperature Effects on the Bystander Signal .............................. 58
1.2.2
In vivo Bystander Effects ......
..........................
.......59
1.2.2.1 Bystander Effects in Pelleted Cell Clusters................................60
1.2.2.2 Bystander Effects in 3-D Tissue Models ................................... 61
1.2.2.3 Bystander Effects in Mice ..................................................... 65
1.2.2.4 LET Dependence of in vivo Bystander Effects ............................. 67
1.3 Thesis Work ...........................................................................
2 Methods and Materials
.. 71
73
2.1 Cell Lines ..............................................................
73
2.2 C ell Irradiation .......................................................................... .. 74
2.3 Co-culture Experiments...........................
..................................... 76
2.4 Radical Scavenger Experiments............................................
....... 77
2.5 Bystander Experiment with 40C Cell Irradiations ...................................... 79
2.6 Micronucleus Assay ............................................................ 80
2.7 y-H 2AX A ssay .....................................
................... .. ...
2.8 Cell Cycle Analysis .........................................
.........
81
............................ 82
2.9 Colony Form ation Assay ................................................................. 82
2.10 Statistical Analysis
.....................................................................................
83
2.11 Experimental Approach Summary...........................................................84
3
Results
86
3.1 M icronuclei Form ation ..............................
....................................... 86
3.1.1
Micronucleus Formation in Bystander DU-145 Cells ........................ 88
3.1.2
Micronucleus Formation in Bystander DU-145 Cells ..........
3.1.3
Temperature Effects on the Production of a Bystander Signal.............. 93
.......... 91
3.2 y-H2AX Foci Expression ...........................................................................
96
3.2.1
Time Course of y-H2AX Induction in Bystander DU-145 Cells............ 97
3.2.2
Time Course of y-H2AX Induction in Bystander AG01552 Cells.........101
3.2.3
y-H2AX Foci Induction in Bystander AGO1522 Cells ....................... 103
3.3 Cell Cycle Analysis of Bystander AG01522 Cells ......................... ... 106
3.4 Survival Fraction of Bystander DU-145 Cells .................................... 108
3.5 Survival Fraction of Bystander AGO1522 Cells ...................................... 109
3.5.1
4
Temperature Effects on the Survival Fraction Bystander Signal.............. 111
3.6 Summary of Experimental Results .....................................................
112
Discussion
114
4.1 LET-Dependence ........... ...
........................................................... 114
4.2 The Nature of the Bystander Signal ..................................
.................. 119
4.3 Different Cell Lines Respond Differently to the Same Signal ..................... 123
5
Summary & Conclusion
127
5.1 Future W ork .......
.................................................................... 129
Bibliography
List of Figures
Figure 1-1. Signaling pathway of the mitogen-activated protein kinase (MAPK) pathway.
Diagram is reproduced from Zhou et al. [88]................................43
Figure 1-2. Schematic of 3-D normal tissue model irradiations with alpha-particle
microbeams. Arrows represent the beam of alpha-particles. For assaying, a perpendicular
cut to the irradiated plane was performed. Diagram is reproduced from Sedelnikova et al.
[75].....
.......................................................................................... 63
Figure 2-1. Alpha particle co-culture irradiation system. (a) schematic of the insert coculture system for alpha particle irradiations. (b) the steel mylar dish on top of the alpha
source, w hich is not visible ........................................................................ 75
Figure 2-2. Chemical reaction of dimethyl sulfoxide (DMSO) with OH ................... 78
Figure 2-3. 2-phenyl-4,4,5,5-tetramethylimidazoline-l-oxyl 3-oxide (PTIO) reaction with
NO .....................................................................................................
78
Figure 3-1: In situ immunofluorescence detection of micronuclei in 2 Gy x-ray irradiated
DU-145. Single arrows: locations of MN, dashed arrow: apoptosis bridge ..............
87
Figure 3-2. In situ flourescene detection of micronuclei in bystander AGO1522 cells cocultured with DU-145 cells. (A) co-cultured with unirradiated tumor cells; (B) cocultured with 1.0 Gy x-ray irradiated tumor cells. Arrows indicate location of MN......88
Figure 3-3. Fraction (%) of bystander DU-145 cells containing micronuclei following 4
hr co-culture with DU-145 cells that had been irradiated with x-rays (3-3A) or alpha
particles (3-3B). The x-ray experiments (3A) were carried out in the presence or absence
of DMSO and PTIO scavengers. The alpha particle experiments (3B) were carried out
with the bystander DU-145 cells present during the irradiation (Before) or added within 5
min after the irradiation (After). Results are the average of at least three independent
experiments. Error bars represent ± 1 SD (** P < 0.01, * P < 0.05 relative to the 0 dose
control) .....................................................
.........................
90
Figure 3-4. Fraction (%) of bystander AGO1522 cells containing micronuclei following 4
hr co-culture with DU-145 cells that had been irradiated with x-rays (3-4A) or alpha
particles (3-4B) in the presence or absence of DMSO and PTIO scavengers. Results are
the average of at least three independent experiments. Error bars represent ± 1 SD. (** P
< 0.01, * P < 0.05 relative to the 0 dose control) ...................................
.......... 92
Figure 3-5. Fraction (%) of bystander AGO 1522 cells containing micronuclei following 4
hr co-culture with DU-145 cells that had been irradiated with alpha particles at 40C, 250C
with immediate co-culture initiation or 250C with a 15 min delay in co-culture initiation.
Data points are the mean ± 1 SD obtained from at least three independent experiments.
(* P < 0.05 for 4oC effect relative to the 25*C effect with x-rays at the same dose).......94
Figure 3-6. Fraction (%) of bystander AGO1522 cells containing micronuclei following 4
hr co-culture with DU-145 cells that had been irradiated with x-rays at 40 C, 250 C with
immediate co-culture initiation or 25 0C with a 7 min delay in co-culture initiation. Data
points are the averages obtained from two independent experiments ........................ 95
Figure 3-7. Fraction (%) of bystander DU-145 cells containing micronuclei following 4
hr co-culture with DU-145 cells that had been irradiated with x-rays at 40C, 250 C with
immediate co-culture initiation or 25 0 C with a 7 min delay in co-culture initiation. Data
points are the mean ± 1 SD obtained from at least three independent experiments. ...... 96
Figure 3-8. In situ immunofluorescence detection of y-H2AX foci induction in DU-145
cells 1.5 h after x-ray irradiation. (A) unirradiated cells; (B) cells directly-irradiated with
5.0 Gy of x-rays............................................ ......................
97
Figure 3-9. Fraction (%) of bystander DU-145 cells containing > 5 y-H2AX foci as a
function of co-culture time with DU-145 cells that had been irradiated with either 2 Gy or
10 Gy of x-rays. Data points presented are from a single experiment, except the 0 dose,
from 8 independent experiments, and 4 h (10Gy), from 3 independent experiments......99
Figure 3-10. Fraction (%) of bystander DU-145 cells containing > 5 y-H2AX foci as a
function of co-culture time with DU-145 cells that had been irradiated with 1.2 Gy of
alpha particles ......................................................
............
99
Figure 3-11. Fraction (%) of DU-145 cells containing > 5 y-H2AX foci as a function of
time after direct irradiation with either 2.0 or 10.0 of x-rays. The data in the irradiated
DU-145 cells were from one experiment .................................................... 100
Figure 3-12. Fraction (%) of DU-145 cells containing > 5 y-H2AX foci as a function of
time after direct irradiation with 6.0 Gy of alpha particles. The data in the irradiated DU145 cells were from one experiment...........................................................
101
Figure 3-13. Fraction (%) of bystander AG01522 cells containing > 5 y-H2AX foci at 18 hr after co-culture with DU-145 cells that had been irradiated with 1.2 Gy of alpha
particles. Results are the average of at least three independent experiments. Error bars
represent ± 1 SD ............................................................ 103
Figure 3-14. Fraction (%) of the bystander AG01522 fibroblast cells showing induction
of y-H2AX foci after 4 hrs co-culture with irradiated DU-145 tumor cells. The x-ray (314A) or alpha particle (3-14B) doses delivered to the DU-145 cells are indicated on the xaxis. X-ray experiments (3-14A) were carried out with and without the addition of
DMSO or PTIO scavengers. Results are the average of at least four independent
experiments. Error bars represent ± 1 SD. (** P < 0.01, * P < 0.05; significance of
scavengers were relative to the same absorbed dose delivered to the directly-targeted cells
in the absence of scavenger) ................................
.............................
105
Figure 3-15. FACS analysis of the bystander AG01522 fibroblast cells after 4 hrs of coculture with unirradiated DU-145 tumor cells ................................................ 106
Figure 3-16. FACS analysis of the bystander AG01522 fibroblast cells after 4 hrs of coculture with 2.0 Gy x-ray irradiated DU-145 tumor cells...................................107
Figure 3-17. Surviving fraction of bystander DU-145 cells following 4 hrs co-culture
with DU-145 cells that had been irradiated with either x-rays (A) or alpha particles (e) at
the doses indicated on the x-axis. Data points are the mean ± 1 SD of three independent
experiments for X-rays and four experiments for alpha particles ........................108
Figure 3-18. Surviving fraction of bystander AG01522 cells following 4 hrs co-culture
with DU-145 cells that had been irradiated with either x-rays (A) or alpha particles (e) at
the doses indicated on the x-axis. Data points are the mean ± 1 SD from four
independent experiments for x-rays and two experiments for alpha particles. (** P <
0.01 relative to the non-irradiated controls) ....................................
................110
Figure 3-19. Surviving fraction of bystander AGO 1522 cells following 24 hrs co-culture
with DU-145 cells that had been irradiated with x-rays at the doses indicated on the xaxis in the presence or absence of DMSO and PTIO scavengers. Data points are the
mean ± 1 SD from four independent experiments with x-rays and five experiments with
scavengers. (** P < 0.01, * P < 0.05 relative to the no-scavenger-added effect with xrays at the sam e dose) ..............................................................................
110
Figure 3-20. Surviving fraction of bystander AGO 1522 cells following 24 hrs coculture with DU-145 cells that had been irradiated with x-rays at (i) 4oC, (ii) 25TC with
immediate co-culture initiation or (iii) 25oC with a 7 min delay in co-culture initiation.
Data points are the mean ± 1 SD from at least four independent experiments. (** P <
0.01, * P < 0.05 for 4oC effect with x-rays relative to the 25TC effect with x-rays at the
sam e dose) .....................................................
..............
.............
112
List of Tables
Table 2-1. Experimental approach......................................................
....... 85
Table 3-1. Fraction (%)of bystander AG01522 cells containing > 5 y-H2AX foci at 5
min-4 hr after co-culture with DU-145 cells that had been irradiated with 2.0 Gy of xrays. Data from two experiments. The control (0 dose) sample was processed after 2 hrs
of co-culture ..............................................................................
..
102
Table 3-2. Summary of bystander effect experiments with three cases (shaded cells) of
bystander effect differences due solely to radiation quality (LET). ' DMSO appears to
block the effect but the differences are not statistically significant ........................ 113
Chapter 1
Introduction
Overview
Radiation biology studies biological effects caused by ionizing radiation. For
decades the field was based on the tenet that only cells that were directly traversed by
radiation would exhibit biological effects. However, in the past decade, there has been an
abundance of evidence provided by several independent research groups indicating
higher numbers of cells exhibiting damage than expected based on the number of cells
traversed by the radiation. This phenomenon has been coined as the "bystander effect".
The bystander effect is the observation that non-irradiated cells near a cell traversed by
radiation also express biological responses such as double strand breaks, micronuclei,
genomic instability, and cell cycle arrest. (For recent reviews see: [1-7])
The bystander effect field is vast and largely contradictory. While the early
literature primarily focused on the bystander phenomenon, in recent years the studies
have evolved to concentrate more on the factors involved in the bystander signal
emission and in the recipient bystander cells. The majority of the published studies on the
have been performed in vitro. A few in vivo bystander studies have been published,
however, the in vivo bystander field is in its infancy. Bystander signaling studies in vitro
are very limited since all effects studied are in monolayers of cells, usually of a single
cell type. Signaling cascades in vivo involve many different types of cells with
specialized functions. For example, some cells in vivo are responsible for sensing damage
and initiating the signal, while others are involved in the amplification and propagation of
the signal. Cells in vivo are also responsible for complex tissue level signal responses. In
vitro studies can only look at isolated components of what is a linear sequence of
complex signaling systems in vivo. The limitation of in vitro studies may explain some of
the contradictions that exist in the bystander literature since each individual study looks
in detail at an "out of context" component of signaling. Nevertheless, bystander effect
studies performed in vitro, do provide preliminary evidence for possible signaling
mechanisms that could occur in vivo. Eventually bystander effect studied must include in
vivo systems in order to understand the true relevance of the bystander effect in radiation
biology.
Bystander effect studies examine either the production of the signal from the
directly-irradiated cells and/or the response in bystander cells that receive the signal.
Some of the key findings from the bystander effect literature are as follows.
(i) The effect is cell line dependent: while both normal and tumor cell lines have
shown to release and/or to respond to a bystander signal, several cell lines do not exhibit
the bystander effect.
(ii) Whether the bystander effect is detected is dependent on the endpoint assayed.
(iii) Most studies have detected serious biological damage in bystander cells that
often eventually leads to increased cell death.
(iv) The additional radiation damage attributed to the bystander effect is most
significant at very low doses, saturates at doses above 0.5 Gy, and is insignificant in high
dose experiments where direct effects dominate.
(v) In some cases, bystander effects can be induced by a single irradiated cell.
(vi) The bystander signaling pathway appears to be a cascade consisting of
reactive oxygen species and growth factors often leading to up and down regulation of
proteins.
(vii) The bystander signal is transmitted either via intracellular communication
channels known as gap junctions or is released into the medium shared by the two cell
populations.
(viii) The bystander effect has been shown in both in vitro and in vivo studies.
Due to the significant increase of detrimental bystander effects in normal cells at
low doses, the bystander field has mainly focused on radiation protection studies.
Available data from atomic bomb survivors has shown a linear dependence between
cancer induction and radiation exposure at high doses. In the low dose region, additional
cancer risk from radiation exposure is difficult to measure due to the already large cancer
background within the human population. Instead, the scientific and policy regulating
community accepted the linear no threshold (LNT) model, which allows for a linear
extrapolation of the cancer risk data at the high doses down to the low doses with
assumptions made to correct for low dose rates and for high-LET exposures. Therefore,
the LNT model states that any amount of radiation exposure increases an individual's risk
of developing cancer. The bystander effect studies have greatly challenged the validity of
the current LNT model. Is the increase in overall biological damage from direct as well as
indirect effects suggesting that cancer risks from radiation in the low-dose region might
be higher than what was previously believed? This question is the driving force in the
majority of the normal cell bystander studies.
In addition to applications in the radiation protection field, the bystander effect is
also a concern in radiation treatment for tumor therapy. Immunotherapy using alpha
particles is currently being studied as a method to irradiate targeted tumor cells. Therapy
using an external beam is used to target the tumor while sparing maximum amount of
normal tissue. It is possible that bystander signals could increase cell damage within the
tumor and can cause additional damage to surrounding healthy tissue. Few in vitro and in
vivo studies have been done on bystander effects in tumor cell lines [8-15]. A few studies
have shown that directly-irradiated tumor cells released a bystander signal to neighboring
bystander tumor cells, which increased the overall biological damage in the tumor
population [10, 11]. A few in vitro studies have shown that directly-irradiated tumor cells
emit a signal that induced greater biological damage in neighboring bystander normal
cells [8, 16].
The bystander effect will only be relevant if the effect occurs in vivo. The two key
applications in humans will then be radiation protection and tumor therapy. Bystander
effects will have to be taken into account when setting protection limits for radiation
exposures. If directly irradiated tumors emit a signal that affects neighboring normal
tissues in vivo, then it might be possible to manipulate this signal to decrease this
additional damage to normal cells. If an in vivo signal from directly-irradiated tumor cells
also increases damage in unirradiated neighboring tumor cells, it would be advantageous
if this signal could also be manipulated to enhance overall tumor cell death.
The work performed in this thesis studied the bystander effect produced by
directly-irradiated prostate tumor cells (DU-145) and transmitted to either bystander DU145 tumor cells or to AG01522 normal human fibroblasts. This study is meant to be an in
vitro representation of the type of possible bystander effects that might occur in vivo
during tumor therapy. Previous results with DU-145 human prostate carcinoma cells with
alpha particle irradiations demonstrated increased micronucleus (MN) formation in the
co-cultured bystander DU-145 cells [10]. This thesis builds upon those results by
including parallel irradiations with 250 kVp x-rays to provide a low-LET comparison to
the alpha particle results. Normal human fibroblasts (AG01522) were added as a second
co-cultured bystander cell line for comparison of a normal cell line effect with the
bystander effects in the DU-145 prostate tumor cells. This matrix of experiments
addressed two questions: (1) Does irradiation of DU-145 prostate carcinoma cells cause
bystander effects in the AG01522 normal human fibroblasts? and (2) Are there LETdependent differences in bystander effects between 250 kVp x-ray and alpha-particle
irradiated DU-145?
Data presented in this thesis will show a unique LET-dependent bystander effect
due to co-culturing of bystander cells with directly-irradiated DU-145 cells. In addition,
comparison of bystander biological damage responses in the form of micronuclei
formation, y-H2AX foci induction, and survival fraction will be presented. DU-145 cells
irradiated with either alpha particles or x-rays will be shown to emit a bystandersignaling cascade that will affect the two bystander cells differently biologically. Radical
scavengers were added to probe the nature of the bystander signal and it was determined
that the signaling pathway from the prostate tumor cells includes reactive oxygen species.
Finally, irradiation of the DU-145 prostate cell line on ice is shown to eliminate the
bystander effect in the normal human fibroblasts, suggesting that the initial bystander
signal from the directly-irradiated DU-145 cells includes protein-like factors.
This introduction section is meant to highlight key published findings within the
bystander literature, and is not meant to be comprehensive. In the sections that follow, the
major types of bystander effect studies will be reviewed. In each section, a few key and
representative literature papers will be described in some detail to illustrate the
complexity of this diverse field. The progression of topics includes:
*
The damage caused to DNA by direct irradiation
*
The types of damage observed in bystander cell in vitro
* In vivo bystander effects
*
LET-dependence of bystander effects
1.1 DNA Damage in Directly-irradiated Cells
As a charged particle travels through matter it deposits its energy largely by
excitations and ionizations of atoms. To quantify this energy deposition, the term linear
energy transfer (LET) is used. LET is defined as the amount of energy transferred per
unit distance in units of keV tm- 1. Typically, a heavy charged particle will lose a small
fraction of its energy with a single electron collision. For example, a proton will lose
0.22% of its energy in an electron collision. The path of a heavy charged particle will be
almost straight with only slight deflections. This type of path is due to the large
differences in mass between the heavy charged particle and the electron. The track
structure of a heavy charged particle traversal through a cell is a linear path of densely-
packed ionizations and excitations. Heavy charged particles are typically considered
high-LET. Unlike heavy charged particles, electrons and positrons will impart a large
portion of their energy in a single electronic collision and will have very torturous paths
with great deflections. Light charged particles and photons are considered to be low LET.
When a cell is traversed by an ionizing particle, DNA is considered to be the critical
target. The type of damage that occurs to the DNA helix is dependent on the LET of the
radiation particle and occurs through two means: direct and indirect action. Direct action
refers to direct ionizations and excitations of the DNA helix and is the dominant damage
process from a high-LET particle. Direct action often leads to clustered lesions, which are
multiple damaged sites within a few turns of the DNA helix. Clustered lesions are
composed mainly of single and double strand breaks, crosslinks, and basic and abasic
damaged sites [17-20]. Clustered lesions are often resistant to DNA damage repair, which
makes a high-LET particle effective at cell killing. In contrast, indirect action refers to the
ionization and excitations of molecules surrounding the DNA, predominately water,
which leads to the production of free radicals. These highly reactive free radicals, which
are atoms or molecules with an unpaired orbital electron, then diffuse and cause chemical
changes to the DNA helix. Indirect action is the dominant damage process from a lowLET particle due to its torturous and sparse track structure. Indirect damage causes
mostly single strand breaks in the DNA helix, which can be accurately repaired.
Therefore, a low-LET particle is not as effective at cell killing as compared to a highLET particle.
1.1.1 DNA Repair
It is critical for a cell to efficiently repair damaged DNA molecules caused by direct
and indirect action from radiation exposures in order to maintain the normal function of
the cell. If a cell is incapable of repair, or incorrectly repairs the DNA helix, the outcomes
that could occur are: (i) a permanent state of dormancy known as senescence; (ii)
programmed cell death known as apoptosis; or (iii) uncontrolled cell division, which can
lead to cancer. While senescence and apoptosis are seen as protective measures by a cell
against cancer, uncontrolled cell division is not. It has been shown that individuals with
DNA repair deficiencies suffer from premature aging related diseases, and have a higher
incidence of cancer [21]. Damaged sites in the DNA helix resulting from radiation
exposure are mainly single and double strand breaks, crosslinks, and basic and abasic
damaged sites [17-20, 22]. A cell is capable of detecting damaged sites on the DNA by
the binding of initial repair proteins, which then begin to recruit other repair proteins to
the site. The mechanism of repair, as well as the associated proteins of the pathway, is
dependent on the type of damage and the cell cycle phase of the cell.
Single strand breaks are relatively simple to repair, and most importantly, the repair
is error free. With only a defect in one strand, the other strand of the double helix will be
used to guide the repair. It is critical for a cell to efficiently repair damaged DNA
molecules caused by direct and indirect action from radiation exposures in order to
maintain the normal function of the cell. The three repair pathways for single strand
breaks are: base excision repair (BER), nucleotide excision repair (NER) and mismatch
repair (MMR). As the name suggests, BER is used for single base damage in the DNA
helix. Essentially, DNA glycosylase is recruited to the location of the base damage and is
used to cut the 13-N glycosidic bond to then create an AP (apurinic or apyrimidinic) site.
The DNA damage is removed on the 5' side of the AP site by AP endonuclease. This
creates a free 3'-OH, where DNA polymerase is then used to replace not only the
damaged nucleotide but also all nucleotides downstream. Finally, DNA ligase seals the
new DNA strand [23]. The NER pathway is similar to the BER pathway but in addition
to recognizing a single lesion, NER repair enzymes can also locate distortions in the
shape of the DNA helix [24]. The numerous proteins in the NER pathway are capable of
detecting a lesion in the DNA. Once the lesion is located, the XPG (Xeroderma
pigmentosum, complementation group G) protein is then used to make a 3' incision and
XPF (Xeroderma pigmentosum, complementation group F) protein is used to make a 5'
incision in the DNA strand, and, in effect, cuts away the lesion. These incisions create a
gap in the DNA strand of roughly 25-30 nucleotides. DNA polymerase is then used to fill
in the gap. Replication protein A (RPA) is used to protect the integrity of the undamaged
strand during NER. DNA ligase is then used to seal the cuts. MMR is predominately used
for mismatched normal bases such as, A/C, which can occur during DNA replication and
recombination. MMR and NER are essentially the same, except different proteins are
employed in MMR than in NER.
Unlike single strand breaks, misrepaired double strand breaks can lead to genomic
rearrangements that result in hazardous outcomes to a cell. Two mechanisms exist to
repair double strand breaks (DSBs): homologous recombination (HR) and nonhomologous end joining (NHEJ) [25, 26]. The HR pathway is only available during the
G2 phase of the cell since the sister chromatid is used as a template during repair. HR
repair is efficient and accurate since the original template of the DNA helix can be
restored. Rad51 is an important protein in HR repair. In contrast, NHEJ repairs DSBs in
non-G2 phased cells so no sister chromatid template is available. NHEJ repair begins by
the recognition of DSBs by DNA-dependent protein kinases, with the Ku complex
(Ku70/Ku80) as a subunit. The Ku complex then recruits MRX complex, which is
composed of Mrel1, Rad50, and Xrs2, followed by a DNA ligase. The bringing of both
ends of the DSB together proceeds with minimal DNA synthesis. Potential for
substantial DNA nucleotide loss on either side of the break can occur and can lead to
important deletions and possible mutations.
When a cell detects DNA damage, cell cycle checkpoints are activated in order to
give the cell time to repair before continuing to divide. Checkpoints occur between the
GU/S boundary and the G2/M boundary. ATM and ATR are the two primary kinases,
which manage checkpoint activation [27]. ATM also targets the induction of p53, which
can induce apoptosis. In addition, p53 can delay the cell in G1/S by activating p21wafl,
which in turn deactivates the CDK1/cyclin B complex. p2lwaflcan also delay cells in G2
[28]. During these checkpoints, cells will undergo DNA repair.
1.1.2 Hypothermic Cell Irradiations
Mice, which received whole-body x-ray irradiations at temperatures of O0C to 5°C,
have shown a decreased sensitivity to radiation damage [29, 30]. Hornsey performed
whole-body irradiations of adult male mice using 250 kVp x-rays [30]. One set of mice
was cooled to temperatures of O0C -I"C where all heart and respiratory movement in the
mice had ceased. These mice were revived 5 min after irradiation. Another set of mice
was irradiated at room temperature. The LD 50 ,3 0 (dose that is lethal to 50% of the
population at 30 days) and weight loss were measured in both groups. The LD 50,30 was
620 rads for mice irradiated at room temperature. Mice that were irradiated at low body
temperature with doses of 900, 1200, and 1500 rads of x-rays exhibited 100% survival at
30 days. While initial weight loss in the cooled mice was seen, within 20 days post cold
treatment weight was regained to normal levels in both the irradiated and unirradiated
populations. Mice irradiated with 900 rads at room temperature lost half of their body
weight by day 20. Therefore, reducing the temperature of the mouse provided the mouse
significant radioresistance to the x-ray exposure.
In vitro, Weiss found no cell growth difference in HeLa cells post x-ray irradiation at
either 1°C or 37°C [31]. Similarly, Smith et al. did not observe any significant survival
fraction differences in mouse bone marrow cells that were directly-irradiated in vitro with
x-rays at 2°C, 25°C, or 37°C [32]. Belli et al. showed that in vitro, HeLa cells did exhibit
survival fraction radioresistance when irradiated with x-rays at 5oC as compared to 37°C
[33], which contradicted the data of Weiss [31]. Production of free radicals due to
radiation effects in cells is independent of temperature (near zero energy for activation of
radical formation). One possible explanation for these results was that temporarily a
hypoxia environment was created due to the hypothermia. To test this, Cater and Weiss
measured the oxygen tension in the spleen and testes in mice that were cooled with the
exact protocol of Hornsey [34]. They measured oxygen tension of 0-2 mm of mercury in
the two organs. Weiss postulated that the added hypothermia radiation protection
observed in vivo must be due to the induced hypoxic environment, and not to any direct
temperature effects.
If cell kill is only dependent on the energy deposition by radiation particles, then
biological damage should be independent of temperature. However, in mammalian
cultured cells, the physical induction of single and double strand breaks following either
low-LET (x-ray) or high-LET irradiations (alpha particles) was independent of
temperature (2°C versus 37°C) [35, 36]. Larsson et al. also showed that in populations of
asynchronous human fibroblasts or in a breast tumor cell line, there was no difference in
MN induction in either cell line while being held at 2°C or 37°C during x-ray irradiation
[37]. Therefore, physical damage in directly irradiated cells is independent of the
temperature at the time of the irradiation. Directly irradiated cells have been shown to
emit a bystander signal. Do hypothermic irradiation conditions affect the ability of an
irradiated cell to emit a bystander signal? This question will be addressed in section
1.2.1.9, Temperature Effects on the Bystander Signal.
1.2 The Bystander Effect
Biological damage to the cell caused by radiation can be shown through activation of
specific proteins, serious chromosomal damage such as micronuclei formation or
chromatin bridges, cell cycle arrest, or cell death via apoptosis or necrosis when damage
cannot be repaired. For decades it was believed that only cells that were directly traversed
by radiation would be damaged due to the direct and indirect actions of the charged
particles with or around the DNA helix. In the past fifteen years, numerous studies from
independent research groups have shown evidence of greater biological damage within
cell cultures than expected based on the number of cells traversed by the radiation, and
the term-bystander effect has been coined. The bystander effect is the observation that
non-irradiated cells (termed bystander cells) near a cell directly traversed by radiation
also express biological responses such as double strand breaks, micronuclei, genomic
instability, cell cycle arrest, and cell death [41-45]. Bystander effects are characterized by
either directly irradiating whole cell populations with broad beam irradiations or with
microbeams. The passage of the signal from the irradiated cells to the bystander cells has
been shown to be either medium-mediated or intercellular gap junction-mediated
communication, or, in some cases, a combination of both.
1.2.1 In vitro Bystander Effect Studies
The earliest bystander studies came out of a surprising observation from radiation
protection studies. Nagasawa and Little were interested in detrimental health effects from
radon exposures in homes. Radon inhalation results in low fluence alpha particle
irradiations to lung tissues, depositing doses in the range of 0.3-0.4 mGy per year [40].
Radon exposures are estimated to cause 10-14% of lung cancer cases each year [39, 46].
Nagasawa and Little irradiated confluent Chinese hamster ovary cells (CHO) growing on
a mylar membrane with 238pu alpha particles such that only 1% of the cells were traversed
by one or more particle as calculated by Poisson statistics. Surprisingly, it was found that
at this exposure, 30% of the cells exhibited sister chromatid exchanges (SCE) [40].
Hypoxanthine guanine phosphoribosyl transferase (HPRT) gene mutations in bystander
CHO cells were also examined. A 71-fold increase in HPRT gene mutations in CHO cells
from an alpha-particle exposure, where it was calculated that only 1-3% of the cell nuclei
were traversed by radiation [47]. The mutation frequency in confluent CHO cells showed
an above non-linear dose response curve at doses below 5 cGy. This again suggested that
at very low dose exposures, a greater number of cells were damaged than the number of
cells traversed by radiation. The mutation frequency was always higher than would have
been expected from traditional linear dose response curves. Nagasawa and Little
postulated that communication in the form of a biological signal must have been
occurring between directly-irradiated cells and neighboring unirradiated cells and the
signal was inducing this higher level of SCEs and HPRT mutants. Immense
1.2.1.1. Gap-junction Mediated Bystander Effects
One possible mode of communication between cells occurs through the involvement
of gap junction-mediated intercellular communication (GJIC). Connexin proteins, which
make up the gap junctions, pass ions, second messengers and small metabolites between
cells that are in contact such as confluent cells. Azzam et al. looked at the role of cell to
cell contact in the bystander effect [48]. Six cell lines were used: normal human-diploid
skin fibroblasts (AG01522), normal human-diploid lung fibroblasts (HLF1), rat liver
epithelial cells WB-F344 and WM-aB 1 which are GJIC- proficient and GJIC-deficient
respectively, and wild type and connexin 43-/- mouse embryo fibroblasts (MEFs). All
cells were grown to confluency on a mylar membrane and irradiated with low fluence
alpha particles from a 238pU source. The gap junction inhibitors used to block intercellular
communication were: lindane (y-isomer of hexa-chloro-cyclo-hexane), DDT
[1,1'bis(pschlorophenyl)-2,2,2-trichloroethane] and dieldrin. The endpoints studied were
induction of p53, p21waf, and micronuclei (MN) formation. It has been well documented
that after direct ionizing radiation, cells with wild-type p53 respond by increasing p53
levels, which in turn cell cycle arrests the cells until damage is repaired or triggers the
cells to undergo programmed cell death (apoptosis) [49]. p2 lwafl is a direct downstream
response protein of p53 that is involved in G cell cycle arrest. Therefore, p21 waf I
induction is an indication that p53 was activated. Micronuclei are formed predominately
from non-rejoined DNA double strand breaks [50]. When both AG01522 and HLF1
fibroblast cell lines were irradiated with 0.16 cGy, a 2-5 fold increase in p21 waf and p53,
respectively, was seen where only 1% of the cells were calculated to be directly traversed
by alpha particles. All three inhibitors of GJIC reduced the induction of p2lwaf1 and p53
to control levels, but only at doses below 10 cGy. The p2 1 wafl and p53 levels remained
increased at doses above 10cGy, where all of the plated cells were directly-irradiated.
This suggested that at doses below 10 cGy, gap junction inhibitors were blocking the
bystander effects. At doses higher than 10 cGy, the biological damage due to the gap
junction dependent bystander effects were insignificant compared to direct effects. In the
rat liver epithelial cells, a p21 waf bystander response was only seen in the gap junction
proficient cell line WF-F344 and no significant p21 waf1expression was seen in the WMaB 1 cells until higher doses and were attributed largely to direct effects. This was
mimicked in the MEF cell line, where p21 wafbystander induction was seen in the wildtype line but not in the connexin43-/- line. The AG01522 cells also showed a 3-fold MN
induction at 1 cGy (7% of cells hit), and the bystander effect was abolished by lindane.
This study, along with others, has shown strong evidence for the dependence of gap
junctions as one mode of communication for the bystander signal [3, 39, 48, 51-58] .
It is still possible that a medium-mediated bystander effect also occurs in gap
junction experiments since the directly-irradiated and bystander cells are growing in the
same medium. Hu et al. grew confluent AG01522 fibroblasts on a mylar layer and
irradiated the cells with low doses of alpha particles from a 241Am source [53]. The
endpoint used was y-H2AX foci induction. y-H2AX, a histone protein that is part of the
chromatin, is phosphorylated at serine 139 at the location of double strand breaks. At a
dose of 1 cGy, the number of nuclei traversed by at least one alpha particle in this study
was 9.2%, but the yield of positive cells with y-H2AX foci was shown to be 42.5%.
Lindane was added prior to irradiation, and the y-H2AX foci induction at 1 cGy was
reduced to 19%. Since gap junctions are too small to pass proteins, it has been
hypothesized that perhaps the bystander effect is an emission of reactive species into the
medium [53]. Reactive oxygen species (ROS) have been shown by a number of
investigators to be part of the medium-mediated bystander-signaling pathway [3, 44, 5153, 59]. Hu et al. added DMSO, a general radical scavenger, prior to the alpha particle
irradiation and saw a reduction to 19% y-H2AX foci induction in the fibroblasts, which
was still a 2-fold increase over control levels. Since neither lindane nor DMSO
completely removed the bystander effect, it was postulated that a bystander signal was
being emitted via gap junctions as well as through the medium.
1.2.1.2. Medium-mediated Bystander Effects
The limitation of low-fluence alpha particle studies is the dependence on statistics to
determine the fraction of cells actually traversed by a particle. To avoid this
methodology, a medium-transfer technique can be employed: population of cells in one
container are irradiated; the medium from the irradiated cells is removed and filtered; the
medium is then added to a population of unirradiated cells, and; the bystander cells are
assayed for biological endpoints. Mothersill and Seymour used the medium-transfer
method to study the bystander effect in four different cell lines: immortal but nontransformed human keratinocytes (HaCAT), immortal but non-transformed fibroblasts
(MSU-1), human prostate carcinoma (PC-3), and human colon carcinoma (SW48) [13,
38]. Directly-irradiated cells, termed donor cells, were irradiated with 60Co gamma rays.
After a one-hour incubation, the medium (termed irradiated cell conditioned medium ICCM) was removed from the donor cells, filtered, and added to unirradiated cells. The
endpoint studied was survival fraction. The HaCAT, PC-3 and SW48 bystander cells
showed survival fraction reductions to 61%, 80% and 10%, respectively, when ICCM
was added from the 5.0 Gy irradiated donor cells to the bystander cells of the same cell
type. The fibroblast cell line, MSU-1, did not show a reduction in survival fraction in the
bystander cells. In addition, when MSU-1 medium was added to bystander HaCAT cells,
no reduction in survival fraction was seen leading to the conclusion that the MSU-1 cell
line does not emit a bystander signal that causes a reduction in survival fraction. To
confirm that the donor cells were the source of the signal, medium alone (no cells) was
irradiated with 5 Gy. No reduction in survival fraction was seen in any of the bystander
cell lines receiving the irradiated medium (no cells). Additionally, the reduction of
survival fraction to 60% in bystander HaCAT cells was dose-independent and plateaued
at this level at all doses studied from 0.5-10 Gy. The bystander effect in HaCAT cells was
dependent on the number of irradiated cells, where the reduction in survival was only
80% for 1000 irradiated cells but 60% for 100,000 irradiated HaCAT cells. In another
experiment, the same total number of HaCAT cells was grown as either single cell
suspensions, microcolonies of 3-4 cells, or in confluency. The survival bystander effect
was greatest from ICCM from the irradiated confluent cells as compared to the single cell
suspensions and the microcolonies. Phorbol myristate acid (PMA), a gap junction
inhibitor, was added to donor HaCAT cells prior to the gamma ray irradiation. Blocking
of the GJIC had no effect on the bystander survival fraction, which led to the conclusion
that in these cells, the bystander signal was independent of GJIC and solely dependent on
molecules emitted into the medium. The lifetime of the bystander signal from HaCAT
cells appeared to be long lived. Adding ICCM from the donor cells 1-60 hours after
irradiation produced the same 60% survival fraction reduction in the bystander HaCAT
cells. Additionally, the bystander HaCAT cells could not be rescued from damage by
removing the ICCM and replacing it with fresh medium 0.1-240 hours later; the survival
fraction reduction was always present. This study, along with others, showed that the
emission of a bystander signal can be cell type-dependent, gap junction-independent,
medium-mediated dependent, long-lived, and dose-independent [38, 43, 60-62].
An additional method used to test medium-mediated bystander effects is a transwell
insert co-culture system [10, 15, 43, 54, 63, 64]. Cells are plated both in a 6-well plate
and in the companion permeable membrane insert, which allows for passage of small
molecules but not cells. The cells on the 6-well plate are irradiated and immediately
afterwards, the unirradiated cells on the inserts are added to the wells. Therefore, the
irradiated and unirradiated cells are co-cultured together, sharing the same medium,
without touching. Yang et al. used the transwell co-cultured system to study the
bystander effect in normal human skin fibroblasts (AGO 1522) using 250 kVp x-rays.
Bystander fibroblasts, which were co-cultured with directly-irradiated fibroblasts, showed
a two-fold induction of serious biological damage: micronuclei and p21 wa. All bystander
results in the fibroblasts were seen to plateau at 0.1 - 2 Gy x-ray dose to the directly-
irradiated fibroblasts indicating once again that the bystander effect was doseindependent. Additionally, the survival fraction of the bystander fibroblasts decreased to
70% at doses to directly-irradiated cells above 0.5 Gy. Yang et al. used the antioxidant
enzymes, copper, zinc - superoxide dismutase (Cu-ZnSOD) and catalase, and showed the
effective removal of the p21 wafland micronuclei bystander induction. These enzymes,
however, did not remove the reduction in survival fraction in the bystander fibroblasts.
The generation of ROS in the cells was measured using DCFH-DA (2',7'dichlorodihydrofluorescein diacetate). Thirty hours after the initiation of transwell coculture, a four-fold increase in ROS production was detected in bystander cells. By 60
hours, the ROS production in the bystander cells returned to background levels. This
experimental protocol only tested for a medium-mediated bystander effect since no cellto-cell contact was present.
1.2.1.3. Microbeam Bystander Effects Studies
The availability of single-particle microbeams has increased the flexibility in
bystander experiments immensely. Briefly, a monolayer of cells is attached to a culture
dish. On the day of the irradiation, cell locations are recorded and stored by an image
analysis computer program. The researcher will then program into the computer the
number of cells that are to be irradiated and with how many particles. The computer then
positions the chosen cell over a highly collimated shuttered beam of acceleratorgenerated particles. A detector is positioned over the cell. The computer opens the shutter
and once the detector records the desired number of particles, the shutter is closed.
Current microbeam technology can allow for the precise delivery of a single particle to
target either the nucleus or the cytoplasm of the cell. By selectively irradiating a fraction
of the cells, recorded populations of directly-irradiated and unirradiated cells are created
within the same culture dish.
Ponnaiya et al. used an alpha particle microbeam to study the bystander effect in
normal human fibroblasts [65]. Two separate populations were stained with either a
nuclear or cytoplasmic dye, mixed in a ratio of 1:1 and plated at a cellular density such
that the mean distance between the cells was 300 gLm. All of the nuclear stained cells
were precisely irradiated with 2, 5 or 25 alpha particles. A 1.3-1.6 fold increase in
micronucleus induction was seen in the bystander fibroblast cells. The bystander response
was the same regardless of the number of alpha particles delivered to the directlyirradiated cells.
It is well known that significant cell cycle delay occurs in cells directly traversed by
radiation in order to allow time for cells to repair DNA damage [66, 67]. Cell cycle delay
has also been documented in bystander cells using both the medium-mediated
methodology as well as broad beam irradiations, which rely on Poisson statistics to
estimate bystander effects [42, 43, 54, 64, 66, 67]. Ponnaiya et al. also monitored the
uptake of BrdU (5-bromo-2-deoxyuridine) in the bystander population. BrdU is
incorporated into newly synthesized DNA and therefore can be used to monitor a cell's
entry into S-phase [65]. All of the nuclear stained cells were irradiated with 1, 5, 8, or 25
alpha particles. Bystander fibroblasts showed a three-fold decrease in BrdU uptake at 24
h post irradiation, and the effect was independent of the number of alpha particles that
traversed the directly- irradiated cells. The bystander cell cycle delay did not persist to 48
h. The G1 cell cycle delay exhibited in these bystander fibroblasts was suggestive of
significant serious biological damage that persists for up to 24 h of co-culture time with
directly-irradiated cells. It was unknown from these data whether the directly-irradiated
cells were continuously emitting a bystander signal for 24h or whether this amount of
time was needed to attempt to repair the biological damage in bystander cells.
Microbeams studies can also be used to irradiate a single cell in an entire population
and look for bystander effects in neighboring cells. Schettino et al. irradiated a single
Chinese hamster cell (V79) with an ultra soft x-ray microbeam and studied the survival
fraction of the neighboring cells [68, 69]. Using the carbon K-shell characteristic x-ray
with a focus spot of less than 1 gtm diameter, doses of 0.1-2.0 Gy were deposited in the
nucleus of a single cell or in all cells. It was determined that the secondary electrons
formed by the x-ray beam would have a track length of no more than 7 nm. Therefore all
the energy deposited by the x-ray microbeam was deposited within the targeted nucleus
and the cytoplasm of the cell. When all the cells in the population were targeted with
doses of 0.1-2.0 Gy, a survival fraction typical of a broad beam low-LET irradiation was
seen. In contrast, a reduction in survival fraction to 90% was seen in bystander V79 cells
that were co-cultured with a single irradiated V79 cell that received a dose of 0.1-2 Gy.
These investigators estimated that for the single irradiated cell, 10-15 additional cells
died due to the bystander medium-mediated effect. Interestingly, at doses below 0.2
mGy, the survival fraction reduction of the V79 cells was the same regardless of whether
a single cell or all-cells irradiation took place in the population. This suggested that at
these very low doses, a single irradiated cell released a bystander signal that induced the
same levels of biological damage as if the entire cell population was irradiated.
Another benefit of the microbeam system is the possibility of automatically
recording the locations and doses delivered to all the cells in the population and to then
revisit these cells at a time later. Schettino et al. recorded the x, y coordinates of the
single irradiated V79 cell and all the additional bystander cells plated, which were up to a
distance of 3 mm away [68]. Three days after the irradiation, a distribution map of the
position of the damaged bystander cells relative to the single irradiated cell was created.
A 10% increase in damage, which correlated to the reduction in survival as mentioned
above, was shown to be present at all distances from the single irradiated cell. The
increase in damage in the bystander cells was the same at all distances up to 3 mm away
and did not change when a single cell was irradiated with either 0.2 or 2.0 Gy x-rays.
This study suggested that: (i) any one given cell was capable of producing a significant
bystander signal; (ii) the bystander signal diffused uniformly throughout the medium and
caused the same bystander damage in bystander cells regardless of distance from the
irradiated cell.
It has been traditionally believed that DNA is the critical target for biological effects
in directly-irradiated cells. However, when alpha particle microbeams were used to
directly irradiate only the cytoplasm of cells, increases in cell killing and genetic
mutations were detected [70]. In the early bystander studies, the statistics used to
calculate the number of cells traversed by an alpha particle only considered the nucleus as
the target and ignored the additional cytoplasmic traversals [39, 40, 48, 71]. Shao et al.
took advantage of the alpha particle microbeam system to directly irradiate either the
cytoplasm or the nucleus of human glioblastoma cells (T98G) and to study the biological
response in a bystander T98G glioblastoma cell population. Two populations of T98G
cells were plated in two separate regions on a plate with a separation distance of 5 mm.
Cells in one of the populations were directly irradiated either in the cytoplasm or in the
nucleus, and 24 h later micronuclei induction was measured in the bystander population.
When a single T98G cell was irradiated in the cytoplasm with either 1 or 10 alpha
particle(s), a 36% increase in micronuclei was detected in the bystander T98G
population, indicating that the bystander effect from cytoplasmic irradiation was
independent of dose delivered to the individual cells. The same increase in MN induction
was seen in bystander T98G cells regardless of whether they were co-cultured with ten
T98G cells that were directly-irradiated with a single alpha particle either to the nucleus
or to the cytoplasm. When either 10 or 100% of T98G cells were irradiated with 1 alpha
particle to the cytoplasm, an equal MN induction was measured in the T98G bystander
cells. To investigate the nature of the bystander signal, PTIO (2-phenyl-4,4,5,5tetramethylimidazoline-1-oxyl 3-oxide), a nitric oxide (NO) scavenger, was added after
the irradiation. PTIO prevented the MN induction in bystander T98G cells that were cocultured with cytoplasmically irradiated T98G cells. This study provided evidence that:
(i) one alpha particle to the cytoplasm of one cell can induce a bystander effect in
neighboring cells; (ii) the magnitude of the bystander effect was independent of the dose
deposited in the cytoplasm of the directly-irradiated cell; (iii) the magnitude of the
bystander effect was independent of the number of cells in the directly-irradiated
population that received doses to the cytoplasm; and (iv) both nuclear and cytoplasmic
irradiations released a ROS-dependent medium-mediated bystander signal.
1.2.1.4. DNA Repair in Bystander Cells
The bystander effect endpoints that have been reported have been cell death,
mutations, and chromosome damage [1, 5, 13, 55]. These effects could have been a result
of a failure to repair the damage in the DNA helix. These studies suggested that
significant double strand breaks (DSBs) have been induced in bystander cells co-cultured
with directly-irradiated cells. It has been shown that one of the early cellular responses to
double strand breaks is the serine-139 phosphorylation of the H2AX histone (y-H2AX)
near the lesion. This phosphorylation can take place within minutes in directly- irradiated
cells and the detection of y-H2AX foci can be used as an early biomarker for DSB
damage [72]. While the exact function of y-H2AX foci formation is unknown, studies
have shown a correlation between DSB locations and y-H2AX foci [73]. Immunostaining
has also shown co-localization of y-H2AX foci and micronuclei in directly-irradiated
human fibroblasts [74].
Several studies have now shown that y-H2AX foci induction was also present in
bystander cells that were co-cultured with directly-irradiated cells [75-77]. Yang et al.
utilized the transwell co-culture method to study DSBs in bystander normal human
fibroblasts (AG01522) that were co-cultured with AG01522 fibroblasts that had been
irradiated with 250 kVp x-rays [43]. The same two-fold y-H2AX foci induction was
detected in the bystander fibroblasts after a 1 h or 24 h co-culture time with the x-rayirradiated fibroblasts. No bystander y-H2AX induction was seen, however, in the
bystander fibroblasts after a 48 h co-culture time. In contrast, in AGO 1522 fibroblasts
directly-irradiated with 0.1 or 2 Gy of x-rays, the y-H2AX foci induction was detected at
1-hour post irradiation, but not at 24 or 48 h. The addition of SOD or catalase removed
the y-H2AX foci induction in the bystander fibroblasts. This study suggested that either
the DSB repair mechanism in bystander cells was slower, or the bystander signal caused
continuous and persistent double strand break damage in bystander cells.
In directly-irradiated cells, it was shown by a number of investigators that yH2AX foci formation was induced within minutes after irradiation [73, 78-80]. The
timing of phosphorylation of the H2AX histone in bystander cells was further studied by
Sokolov et al. [75, 77]. The study used two experimental methodologies: microbeams
and medium transfer. In the microbeam experiments, normal human lung fibroblasts
(W138) were plated on a dish and target cells were irradiated with 2 or 20 alpha particles.
In the directly-irradiated W138 cells, y-H2AX foci induction was seen as early as 30 min
post irradiation, and the effect increased linearly with the number of alpha particle
traversals per cell. By 18 and 48 h, the y-H2AX foci induction was at background levels
in the directly-irradiated cells. A 2-fold y-H2AX foci induction was detected in bystander
W138 cells that were co-cultured for 18 and 48 h with directly alpha-particle irradiation
of W138 cells. No increase in y-H2AX foci was detected in the bystander W138 cells that
were co-cultured for only 30 min with directly irradiated W138 cells. In the mediumtransfer experiments, W138 cells were irradiated with doses of 0.2, 0.6 or 2 Gy using
137Cs
gamma-rays and the medium was removed, filtered, and added to unirradiated
W138 cells. In the cells directly irradiated with gamma-rays, y-H2AX foci induction was
detected as early as 30 min. The y-H2AX foci induction in the directly irradiated cells
persisted but was only slightly elevated at 18 and 48 hrs. A 2-fold y-H2AX foci induction
was detected in the bystander cells at 18 h and 48 h after receiving ICCM. No increase of
y-H2AX foci was detected in the bystander cells at 30 min and 4 h after receiving ICCM.
The addition of the NO scavenger PTIO effectively blocked the W138 bystander y-H2AX
foci induction caused by medium transferred from gamma-ray irradiated W138 cells. In
addition, co-localization of Mrel 1, Rad50, ATM, 53pbl, and Nbsl with y-H2AX foci
was found in the bystander cells from both the alpha particle or gamma-ray irradiation
experiments. These proteins are associated with DNA repair. The data therefore showed
that active repair was occurring in the bystander cells. The data presented in this study
suggested that: (i) repair mechanisms were different in bystander and directly-irradiated
cells; (ii) repair persisted for much longer in bystander cells than in directly-irradiated
cells (also shown by the Yang et al. studies); and (iii) the bystander y-H2AX foci
induction was NO-dependent.
The Replication Repair Protein A (RPA) has been shown to be a vital component
of several single strand break (SSB) repair pathways: nucleotide excision repair, base
excision repair (BER) and recombination repair. Balajee et al. studied the mediummediated bystander effect induction of RPA and MN in two normal human lung
fibroblast lines, W138 and MRC5 [81]. Using the medium transfer method, medium was
removed and filtered 1-hour after a 5 or 10 Gy
137 Cs
gamma-ray irradiation of confluent
fibroblast monolayers. The filtered medium was then added to an unirradiated fibroblast
bystander population. In MRC5 bystander cells, a 2-3 fold RPA bystander induction was
seen as soon as 30 min and persisted to 4 h after addition of conditioned medium from
irradiated W138 cells. RPA levels increased at 30 min and remained elevated up to 6
hours in the directly-irradiated W138 cells. AP endonuclease (APE), a vital enzyme in
the BER pathway, was also assayed in the bystander MRC5 cells. A 2.5- fold increase in
APE was detected in bystander MRC5 fibroblasts that received conditioned medium from
10 Gy gamma-ray irradiated MRC5 fibroblasts. Finally, a 3-fold MN induction was
detected in the bystander fibroblasts that received conditioned medium from gamma-ray
irradiated fibroblasts. This study suggested that: (i) increase of RPA and APE foci
formations were the result of repair of single strand breaks and oxidized base damage in
bystander cells; (ii) BER played a role in the bystander response; (iii) the increase of MN
suggested that double strand breaks were formed in the bystander cells, but; (iv) the
DSBs were not efficiently repaired in the bystander cells.
The growing evidence for DNA double strand damage in bystander cells led
several research groups to study whether cells that have deficiencies in DNA repair
would exhibit larger bystander effects. Little et al. studied the participation of the NHEJ
repair pathway in bystander cells for induction of chromosome aberrations [71]. This
study employed four mouse knockout cell lines that had NHEJ repair deficiencies: Xrcc5
(Ku80-/-), G22p1 (Hu70-/-), and Prkdc (DNA-Pcks-/-). For NHEJ repair-proficient
controls, wild type S or SV cells were used. All cells were grown on a mylar layer to
confluency and irradiated with doses of 0.17 -1.7 cGy, where 2-15% of the cells would
be expected to be directly traversed by an alpha particle. The bystander endpoint studied
was chromosomal aberrations, which included chromatid- and chromosome- type breaks
and exchanges, and rings and dicentrics. At 0.33 cGy (3.3% of the nuclei irradiated),
9.3% and 9.4% of the wild type S or SV cells showed an increase in chromosomal
aberrations, respectively, which represented a 4-fold increase over background levels. In
comparison, at 0.33 cGy, 55% and 45% of the Xrcc5 and G22p1 cells showed an increase
in chromosomal aberrations, which represented a 40-50 fold increase over background.
Similarly, at 0.17 cGy (2% of the nuclei traversed), 20% of Prkdc cells showed an
increase in chromosomal aberrations, which represented a 20-fold increase over
background. It should be noted that the background control level of chromosome
aberrations in the repair-deficient cell lines was higher than in the wild type cell lines.
This study confirmed that double strand breaks occurred in the bystander cells. More
specifically, the involvement of NHEJ repair of double strand breaks in bystander cells
was crucial and deficiency in this pathway resulted in significant sensitization of the cells
to bystander effects.
Bystander survival fraction studies have suggested that the cell losses in the
bystander population must have been those cells in which the DNA damage was not
properly repaired. Therefore bystander cells with repair deficiencies should exhibit much
greater decreases in survival fraction. Mothersill et al. used their standard mediumtransfer method to irradiate five repair-deficient cell lines in the MMR and NJEH
pathways [82]. Each repair-deficient cell line had a parallel repair-proficient cell line as a
control. Repair-proficient or deficient-cell lines were irradiated with gamma ray doses of
0.5 - 5.0 Gy from a 60 Co source. One hour after irradiation, ICCM was removed, filtered,
and added to unirradiated repair proficient or deficient cells. All of the bystander repairdeficient cell lines receiving ICCM from directly-irradiated repair-deficient cell lines
exhibited a significant decrease in survival fraction that in most cases was larger than
what was seen in the parallel repair-proficient cell line experiments. Some of the repairproficient cell lines did not show a bystander effect. Additional bystander experiments
were performed with repair-deficient cell lines receiving ICCM from repair-proficient
cell lines, and vice versa. The ICCM from repair-deficient cells produced more
clonogenic cell death in the corresponding bystander repair-proficient cells. Therefore,
this study suggested that: (i) the bystander signal from repair-deficient cell lines was
more toxic than the bystander signal from repair-proficient cells; (ii) bystander effects
were increased in repair-deficient cells receiving ICCM from repair-proficient or
deficient cells as compared to their repair-proficient counterparts. These investigators
concluded that DNA repair mechanisms were crucial to reverse the effects of the
bystander signal in unirradiated cells.
Double strand breaks that are not efficiently repaired can lead to serious
chromosome damage, mutations and genomic instability in directly-irradiated cells.
Mutations and genomic instability have also been shown in bystander studies [45, 55].
Initial work from Nagasawa and Little showed that a low fluence alpha particle
irradiation of wild-type CHO cells induced HPRT mutations to a level that was
significantly above the estimated number of irradiated cells. The additional mutants were
attributed to bystander effects [47]. Nagasawa and Little followed up on these initial
results by comparing the induction of HPRT mutations in the CHO cells as well as in the
xrs-5 mutant cell line [47, 83, 84]. Xrs-5 cells are deficient in double strand break repair
and have been shown to be radiosensitive. In the wild-type CHO cells, after an irradiation
of 0.83 cGy where less than 5%of the nuclei were traversed by an alpha particle, a 4.6fold increase in HPRT mutants was detected. In comparison, in the xrs-5 cells, after an
irradiation of 0.83 cGy, a 50-fold increase in HPRT mutants was detected. In addition,
PCR analysis showed that, in the bystander CHO cells, 90% of the gene deletions were
point mutations [85]. In contrast, the xrs-5 cells had large-scale mutations and partial
gene deletions at these low doses. This study suggested that bystander cells that had a
compromised DNA repair capability exhibited much greater levels of mutations largely
due to the non-repaired DSB.
While p53 has been implicated in bystander mutagenesis, the role of p53 in
bystander signaling has also been studied. Zhang et al. used three human lymphoblastoid
cell lines from the same donor that differed in p53 status: TK6 (wild type), WTK
(mutant p53), and NH32 (null p53) [86]. In addition, NHEJ and HR repair pathways were
compromised in these cells by RNA interference techniques: DNA-PKcs and Rad 54
were suppressed in the TK6 cells. The directly-irradiated cells received a gamma-ray
dose of 2 Gy from a 137Cs source. The transwell co-culture method was used to study the
bystander effect and the biological endpoint used was the increase of mutant fraction at
the TK locus in the bystander cells. A similar 2-fold TK locus mutation increase was seen
in the three types of bystander lymphoblastoid cells. This suggested that in this cell line,
the initiation and reception of the bystander signal was independent of p53 status.
Subsequently, all further work was done only in TK6 cells. The bystander knockdown
DNA-Pkcs cells showed a 3-fold increase in the mutant fraction of the TK locus as
compared to the 2-fold bystander response in DNA-Pkcs proficient cells. The bystander
cells in which Rad54 was suppressed showed no significant increase in bystander
mutagenesis induction as compared to the bystander Rad54 proficient cells. This study
suggested that in these cells: (i) p53 was not necessary for expression of a bystander
effect; (ii) bystander cells did not depend on the HR pathway; (iii) the NHEJ repair
pathway was important in bystander cells, and therefore; (iv) significant repair of double
strand breaks was occurring in bystander cells.
Cell cycle delays are associated with DNA repair. Several research groups have
studied cell cycle delays in bystander cells [39, 42, 43, 54, 87]. Downstream of p53 is the
p2lwafl protein, which induces a G1/S cell cycle arrest. Studies by Azzam et al. have
shown that in confluent monolayers of five different fibroblasts, irradiation with 1 cGy of
alpha particles (9% of the cells hit), caused significant elevation of p2 1 waf and p53 up to
16 hours post irradiation. In addition, a significant decrease in the cell cycle regulated
proteins CDC2, CCNB 1 and RAD51 was also detected [39]. These effects could not be
attributed only to the small population of directly-hit fibroblasts and therefore suggested
that the additional cell cycle effects occurred due to a bystander effect. Ponnaiya et al.
showed a decrease in BrdU labeled bystander fibroblasts that were co-cultured with
alpha-particle irradiated fibroblasts [65]. Since BrdU labeling is a measure of cells
entering S phase, the study showed that the bystander cells were in a G1 cell cycle arrest
24 hrs after the initiation of co-culture with irradiated cells.
1.2.1.5. Apoptosis in Bystander Cells
Many studies have shown that the mitogen-activated protein kinase (MAPK)
pathway, which is diagramed in Figure 1-1, is linked to proliferation, senescence, and
apoptosis. The MAPK pathway has been shown to be initiated by the binding of the cell
membrane receptors by either the transforming growth factor a (TGF- a), tumor necrosis
factor p (TNF- P), insulin-like growth factor binding protein 3 (IGFBP-3) gene, IL-1, and
interleukin-8 (IL-8) [88]. ROS have also been shown to trigger the MAPK signaling
cascade and nitric oxide is known to regulate the induction of IL-8 [88]. The MAPK
superfamily includes the extracellular signal-regulated kinase 1/2 (ERK 1/2), c-Jun NH 2-
terminal kinase (JNK) and p38. The activation of the MAPK pathway kinases ERK 1/2,
JNK, and p38 will lead to the activation of the cyclooxygenase-2 (COX-2) signaling
pathway. COX-2 has been shown to cause cellular inflammatory responses, mutations,
transformations, and genomic instability. COX-2 can be stimulated by the cytokines
TGF- a and TNF- 1, IGFBP-3 gene, and IL-8.
I"
h1 * -
r'^• 7 o
I -r%
t
4
II
I[
Il
Cell Membrane
MKK4/
JNK
I
7
wt-LPI3 kinase
S
1KB-NF-kB
MEKI/2
MKK3.S
AkT/PKB
ERKWI/2
NF-kB
!L
I COX-2 I
Inflammation, Mutation, Transformation, Genomic Instability
Figure 1-1. Signaling pathway of the mitogen-activated protein kinase (MAPK)
pathway. Diagram is reproduced from Zhou et al. [88].
A major mode of death in bystander cells has been shown to occur via the apoptosis
pathway. In directly-irradiated cells, BCL2 gene expression was reported to be induced
30-60 min post irradiation. BCL2 is a key gene in apoptosis regulation. Caspase 9 is also
activated during apoptosis. There is also increasing evidence that apoptosis-inducing
agents cause an increase in mitochondrial mass per cell and that the mass increase is ROS
dependent [89]. Mothersill et al. used human keratinocytes immortalized with the HPV
virus (HPV-G) cells and the medium-transfer method to look for expression of apoptosisrelated signaling effects in bystander HPV-G cells [89]. Briefly, HPV-G cells were
irradiated with 0.005, 0.05 and 5.0 Gy of gamma-rays from a 60Co source. One hour later,
the medium was removed, filtered and added to unirradiated HPV-G cells. The endpoints
studied in the bystander HPV-G cells were mitochondrial mass per cell (using a
fluorescent dye), BCL2 expression, and cell survival. In the bystander HPV-G cells, a 2fold increase in mitochondrial mass per cell was detected when the ICCM was added
from HPV-G cells that had received 0.005 and 0.05 Gy. No increase in mitochondrial
mass per cell in bystander cells was seen when ICCM was added from cells that received
5 Gy. In addition, the bystander cells that were incubated with ICCM from 0.005 or 0.05
Gy exposures showed a 3-fold change in the number of cells in which the cells the
distribution of mitochondria was perinuclear rather than cytoplasmic. This change in
distribution of mitrochondia was not detected when ICCM from cells receiving 5.0 Gy
exposures was used. Since ROS have been implicated in the induction of increases in
mitochondrial mass per cell, a radical scavenger, NAC (N-acetylcysteine), was added to
the bystander cells. NAC abolished the increase in mitochondrial mass in bystander cells.
Interestingly, only HPV-G bystander cells that received ICCM from a 5.0 Gy exposure
exhibited a 4-fold increase in BCL2 expression and no increase in expression was seen at
the other doses. Adding ICCM to bystander HPV-G cells reduced the survival fraction to
-70%. Addition of a caspase 9 inhibitor blocked cell death in the bystander cells only for
treatment with 0.005 Gy ICCM but not for 5.0 Gy ICCM. This study suggested that in
these bystander cells, apoptosis occurred through a caspase-dependent pathway after
exposure to <0.5 Gy ICCM, and occurred through a caspase-independent pathway after
exposure to 5.0 Gy ICCM. However, the apoptosis pathway was induced in the bystander
HPV-G cells after all exposures to ICCM and could be the cause of cell death seen in the
bystander cells.
Increases in intracellular calcium [Ca 2+]have been shown to cause mitochondrial
production of ROS and apoptosis [59]. The Mothersill group expanded their studies to
test if the MAPK pathway was involved in the cell loss exhibited in HPV-G bystander
cells [59]. Briefly, ICCM was removed from HPV-G cells one hour after 0.5 Gy gammaray irradiation and added to bystander HPV-G cells. The endpoints assayed in the
bystander cells were the induction of ERK, JNK, and p38. Intracellular calcium levels
and mitochondrial potential in the bystander cells were also assayed. Apoptotic bystander
cells were also scored 24 hrs after the addition of ICCM. Bystander HPV-G cells
receiving ICCM showed an 8-fold increase in apoptosis induction. Additionally, the
HPV-G bystander cells showed an increase in activation of ERK 1/2 and JNK, but not
p38. Radiometric measurement of calcium showed a significant increase in bystander
HPV-G cells. The addition of a voltage-dependent Ca 2+ channel blocker abolished the
bystander effect. In contrast, depletion of intracellular calcium stores only partially
removed the increase of Ca2+ levels in bystander cells. These data suggest that Ca 2+
released via the voltage channels played a larger role in the bystander effect with a slight
involvement of the calcium stores. A decrease in mitochondrial membrane potential was
also seen in bystander HPV-G cells, and the decrease was effectively abolished with the
addition of ROS scavengers. This study suggested that ICCM induced increases of
intracellular Ca2+ to flood the mitochondria, which then caused the loss of membrane
potential in the bystander cells. Further, the production of ROS then led to the activation
of the MAPK pathway. The activation of the MAPK pathway resulted in apoptosis as the
mode of death in these bystander cells. Cell death was detected via the reduction of
survival fraction [89].
ROS/NO can be continuously produced in a feedback mechanism in cells via the
plasma membrane - bound NAD(P)H oxidases [90]. NAD(P)H oxidases, which are
localized in lipid rafts, has been implicated as a potential source for persistent ROS
presence in bystander cells [1, 59]. Moreover, the plasma membrane contains lipid rafts
enriched in cholesterols, and these lipid rafts have been shown to contribute to bystander
effects [14, 16]. Oxidative products, such as H20 2, that are caused by persistent oxidative
stress have been shown to induce y-H2AX foci induction in cells as well as single strand
breaks and oxidative base damage through a cascade effect [77, 81]. NO in high
concentrations has been shown to cause micronucleus formation, mutation and DNA
damage [15]. Azzam et al. inhibited the NADP(H) oxidases in the AG01522 cells and
found that no increase in p21wafl or p53 was detected in the bystander cells [90]. Their
data further suggested that ROS production originating from NADP(H) oxidases was a
key component of the bystander-signaling pathway. Azzam et al. had also shown
significant accumulation of p21 waf, p53, ERK1/2, JNK, and p38 in their bystander
studies using AG01522 fibroblasts [90]. Radical scavengers blocked the increases in the
MAPK pathway proteins as well as the p21wafland p53 increases. It was not possible to
determine, however, whether the activation of the NADP(H) oxidases was correlated to
the MAPK activation seen in the bystander fibroblasts.
Zhou et al. looked for inductions of COX-2 and MAPK family proteins in bystander
normal human lung fibroblasts (NHLF) [88]. NHLF cells were plated on a grid-strip
mylar layer and irradiated with 0.5 Gy alpha particles. Cells that were to be directlyirradiated were plated on the mylar layer and the bystander cells were plated on the mylar
layer with steel strips underneath it to block the alpha particles. A 2-fold increase in
HPRT mutations was observed in the bystander HNLF cells that were co-cultured with
directly-irradiated HNLF cells. Additionally, a 3-fold increase of COX-2 was seen in
bystander HNLF cells. Interestingly, IGFBP-3 expression was 7-fold lower in the
bystander cells, which suggested that another factor binds to the cell membrane receptor
to induce the MAPK and COX-2 signaling pathways. ERKI/2 was up-regulated in
bystander cells after 4 and 16 hours of co-culture time. The addition of a COX-2 or ERK2 inhibitor effectively abolished the increase of bystander HPRT-mutant HNLF cells.
This study confirmed that MAPK signaling, COX-2 signaling, and associated genes and
proteins were involved in the in bystander effect responses in normal cells.
Activation of MAPK signaling has also been shown in bystander effect studies
using tumor cells. Facoetii et al. collected medium from human glioblastoma T98G cells
directly irradiated with gamma-rays and transferred the medium to an unirradiated
population of T98G cells [91]. Immunostaining for the IL-8 receptor CXCR1 was done
20 h after the addition of the irradiated medium to the bystander cells. CXCR1 expression
was increased in bystander cells receiving conditioned medium from directly-irradiated
cells after 0.5 or 1 Gy. However, no CXCR1 induction was seen in the bystander tumor
cells that received irradiation conditioned medium from tumor cells irradiated with 2.0 or
5.0 Gy. To compliment the CXCR1 results, this group measured significant releases of
IL-8 into the medium from directly-irradiated T98G cells 20 hours after irradiation. This
study provided evidence that directly-irradiated tumor cells emitted the IL-8 cytokine into
the medium after low-dose irradiations, which activated the CXCR1 receptors. Since IL-8
binds to the membrane receptor that initiates the MAPK and COX-2 pathways, it can be
suggested that these two pathways are part of the bystander response in tumor cells.
It has been shown that significant doses of radiation to a prostate tumor cell line
(DU-145) initiated the MAPK pathway through EGF and TGF- a signaling [9]. Hagan et
al. used the medium transfer method to study the MAPK activation in bystander DU-145
cells. A 6-fold increase in MAPK activity was detected in the bystander DU-145 cells
180 min after the addition of ICCM from a 2.0 Gy gamma-ray exposure. The addition of
a TGF- a antibody effectively abolished the increase of MAPK activity in the bystander
DU-145 cells. This study suggested that TGF- a was released into the medium by
gamma-ray irradiated DU-145 cells. This release caused the activation of the MAPK
pathway in the bystander DU-145 cells.
1.2.1.6. The Adaptive Response and Bystander Effects
The adaptive response is the ability of a small low dose treatment to have a
biological protective effect in cells to a subsequent high dose treatment [2]. The extent of
damage seen in unirradiated bystander cells is comparable to the level of biological
damage exhibited in cells directly receiving a low dose exposure. Therefore, the
bystander signals released into the medium could be viewed as a priming dose for
adaptive response experiments. A few groups have addressed the question of whether
bystander effects can cause adaptive responses in cells. Iyer et al. used normal human
diploid lung fibroblasts (HLF-1) cells, which have been shown by this group and others
to exhibit both medium- and gap junction-mediated bystander effects, to look for an
adaptive response caused by the bystander signal [48, 60, 92]. Confluent HLF-1 cells
were irradiated with 1 cGy of alpha particles, and 1 hr later, the medium was removed
and placed on unirradiated cells. Cells that were treated with the ICCM were then
irradiated with either a dose of 10 or 19 cGy. For comparison, control groups of cells
received only 10 or19 cGy. Survival fraction of the directly-irradiated and the bystander
plus radiation exposure cells was measured. The cells directly irradiated with 10 or 19
cGy of alpha particles exhibited a reduction in survival to 85% and 65%, respectively.
Cells that had received the 1-cGy ICCM, and 4 hrs later received a challenge dose of 10
or 19 cGy, exhibited a reduction of survival to only 95% and 90%, respectively. The
bystander signal had a protective effect in the HLF-1 cells to a subsequent high dose
exposure. This experiment demonstrated that factors emitted into the medium that cause
bystander effects could be viewed as equivalent to a small priming dose in an adaptive
response experiments.
In another study, Mitchell et al. studied whether the bystander effect could be
reduced if bystander cells received a low priming dose [61]. Chinese Hamster Ovary
(CHO) cells were grown to confluency on a mylar dish constructed with steel grids to
create shielded areas during an alpha particle exposure. CHO cells were irradiated with a
2 cGy priming dose using a 250 kVp x-ray source. Five hours later, the mylar layer was
irradiated with a 5 Gy dose of alpha particles. The CHO cells plated on top of the steel
grid did not receive an alpha-particle dose but, instead, became the bystander population.
For comparison, the traditional bystander experiment was performed with CHO cells
plated on the grided-mylar dish, by irradiation with a 5 Gy dose of alpha particles. For
either bystander or adaptive response experiments, the cells were co-cultured for 24 or 48
h, after which the bystander cells were removed and further processed for the survival
fraction assay. A reduction in survival fraction to 77% was exhibited in bystander CHO
cells co-cultured with 5 Gy alpha particle irradiated CHO cells. When the bystander tobe-cells were first irradiated with a 2 cGy priming dose of x-rays, and then co-cultured
with the 5 Gy alpha-particle irradiated CHO cells, the survival fraction at either time 24
or 48 h was 96%. This study suggested that a priming dose caused the CHO cells to be
less sensitive to the bystander signals emitted from directly-irradiated CHO cells.
Therefore, there was an adaptive response against the bystander effect.
1.2.1.7. LET Dependence of the Bystander Effect in Vitro
LET dependence of biological damage in directly-irradiated cells has been studied
extensively and is well understood. The numerous approaches and variations to bystander
effect studies have made it very difficult to come to any solid conclusions regarding the
LET dependency of the in vitro bystander effect. A few studies have kept all
experimental approaches the same, and changed only the radiation type to test for LET
dependence of the bystander effect. Interestingly, one of the first reports on the bystander
effect had also shown an LET dependence of the bystander signal. Hickman et al.
irradiated confluent monolayers of rat lung epithelial cells with either high-LET alpha
particles or with low-LET x-rays [46]. At a dose of 0.6 cGy, Poisson statistics indicated
that only 0.54% of the cells were directly traversed by an alpha particle. Considerably
higher p53 induction was measured in the cell population as compared to the number of
cells actually traversed by radiation. This additional induction was attributed to additional
damage created by communication between irradiated and unirradiated cells. In contrast,
no increase in p53 induction was measured in the cells at doses below 5 cGy of x-rays. It
can be postulated that there was a threshold effect for the x-ray exposures for induction of
p53. The data suggested that no bystander signaling occurred from cellular exposure of 5
cGy of x-rays. This was in contrast to the alpha particle exposure of 0.6 cGy where
bystander effects were seen.
Mothersill et al. have shown a stable medium-mediated bystander response in
HPV-G cells [59, 93]. One hour after HPV-G cells were irradiated with a 60Co source, the
ICCM was removed, filtered and added to an unirradiated population HPV-G cells. The
bystander HPV-G cells exhibited significant survival fraction reduction as well as
increases in intracellular calcium when co-cultured with ICCM prepared with doses of 2
mGy - 1 Gy. Liu et al. used the exact same protocol but, instead, directly-irradiated with
neutrons generated by the 7Li(p,n) 7Li reaction. This neutron source also exposed the cell
populations to a 1 mGy gamma-ray dose [94]. When the HPV-G cells received ICCM
from gamma-ray doses equal to or below 2 mGy, no significant reduction of survival
fraction was seen, nor were there any increases in intracellular Ca2+. The data suggested
that in these cells, there was a dose threshold for medium-mediated bystander effects
from gamma-ray irradiations. Also, the 1 mGy gamma-ray component of the neutron
dose should not induce bystander effects. No survival fraction reduction was seen at any
of the neutron doses studied: 1.3 mGy to 1.0 Gy. These data suggested that the bystander
signal was different with a neutron dose than that from a gamma-ray dose to the HPV-G
cells.
HPV-G bystander cells that received gamma-ray irradiated ICCM had also
previously shown significant increases in BCL2 expression and decreases in
mitochondrial membrane potential [89]. These bystander effects in the HPV-G cells were
shown to be ROS dependent [89]. Lyng et al. used the same HPV-G cells but studied the
medium-mediated bystander effects with a proton microbeam [93]. Two populations of
HPV-G cells were plated 5 mm apart. One HPV-G population was irradiated with either
1 or 10 protons and the other population was the bystander cells. A 7-fold increase in
BCL2 expression, a 7-fold apoptosis induction, a 3-fold increase in ROS production, and
a significant reduction of mitochondrial membrane potential was detected in the
bystander HPV-G cells. No significant difference in the bystander effect was detected
following direct irradiation with 1 or 10 protons. More over, no significant difference was
seen between the effects measured in the HPV-G cells directly irradiated with protons
and the effects in the bystander HPV-G cells. This suggested that the damage caused in
the HPV-G cells due to the bystander signal was of the same magnitude as in the cells
being directly-irradiated with 1 or 10 protons. The proton microbeam with an LET of 13
keV/lpm gave the same degree of bystander effects as the gamma-ray bystander studies.
The data suggested that, at the LET values studied, the bystander effect in the HPV-G
cells was LET independent.
Using the transwell co-culture system, Shao et al. studied the medium-mediated
bystander effect in human salivary gland neoplastic (HSG) cells using a carbon beam,
with LETs of 13 keV/gtm and 100 keV/gpm [15]. A 2-fold increase in MN was seen in
bystander HSG cells that were co-cultured with 13 keV/plm carbon-irradiated HSG cells.
In comparison 2.5-fold increase in MN was seen in bystander HSG cells that were co-
cultured with 100 keV/Jgm carbon-irradiated HSG cells. PTIO blocked the MN induction
in the bystander HSG cells co-culture with 100 keV/lpm carbon-irradiated HSG cells. NO
was measured in the co-cultured medium and was found to be in higher concentration
from the high-LET carbon irradiation compared to the low- LET carbon irradiation. The
concentration of nitrate, a product of NO, was found to be around 0.5 - 1.5 gM in the
medium following irradiation. At this concentration, it was estimated that the half-life of
NO in the co-culture medium would possibly be a few minutes. With in this time frame,
NO had enough time to diffuse to the bystander cells and to activate the NADH oxidase,
and eventually the MAPK pathway. While the data reported could be used to conclude
that the MN formation the concentration of NO in bystander cells was LET dependent,
the authors did not perform statistical significance to validate the conclusions. The higher
LET carbon beam had induced a slightly larger bystander effect.
LET dependence of the bystander effects in AG01522 fibroblasts was tested using
microbeams of 380 keV/pm 20Ne and 1260 keV/m 40Ar particles [95]. AG01522
fibroblasts were grown to confluency on a mylar layer. Microbeams were used to either
directly target a total of 1-121 fibroblasts with 1 particle or, alternatively 49 cells were
irradiated with 1 to 4 particles. The same 2-fold MN induction was detected in bystander
AG01522 cells co-cultured with 1-121 directly-irradiated AGO 1522 cells with either the
neon or argon particles. When only 1 cell was targeted in the population, a 1.4 fold
increase in MN was detected in the bystander population. This suggested that even a
single random cell irradiated with one neon or argon particle was capable of emitting a
bystander signal and producing a bystander response. It was calculated that one irradiated
cell induced MN formation in 3000 unirradiated cells. A similar 2-fold increase in MN
expression in bystander AG01522 cells was seen when co-cultured with 49 directlyirradiated fibroblasts that received 1-4 neon or argon particles. When DMSO was added
to the medium, the MN induction was reduced in the bystander cells that were cocultured with directly irradiated cells with either radiation particles. When DMSO and
PMA, a GJIC inhibitor, were added together, the MN induction was abolished in the
bystander cells that were co-cultured with directly irradiated cells with either radiation
particles. The data suggested that at this very high LET range, the bystander effect was
independent of LET in the AG01522 fibroblasts. It is difficult to come to an LET
dependence conclusion since both particles were in the high LET range. The complete
inhibition of the bystander signal by DMSO and PMA suggested that GJIC was more
critical for passage of the signal than medium-mediated communication.
Studies using a true spectrum of radiation qualities to examine the LETdependence of the bystander effect are limited. Boyd et al. used either the UVW human
glioma cells or the EJ138 human bladder carcinoma cell line, which were both
transfected to transport MIBG (metaiodobenzylguanidine). A medium-mediated
bystander effect was studied using either a 60Co gamma-ray irradiation or a MIBG
radiophamaceuticals tagged with
131I-
(beta emitter),
123I
(auger emitter), or 211At (alpha
emitter) [96]. Two hrs following irradiation, the medium was removed, filtered and added
to unirradiated cells of the same type. The bystander effect was measured by assaying for
survival fraction in the unirradiated cells. For tumor cells irradiated with external beam
60Co
gamma rays, the decrease in survival of the bystander cells plateaued at - 70%, a
phenomenon that has been reported by others [13, 97, 98]. The survival fraction in
bystander cells exposed to medium from 13LI-MIBG-irradiated tumor cells showed no
plateau, and continued to decrease to 20%. The survival fraction in tumor cells exposed
to medium from either 123I-labeled or 211At-labeled MIBG showed a U-shaped behavior,
decreasing at low exposures, beginning to plateau at ~ 50-70% survival fraction, but then
increasing as the exposure level to the directly-irradiated cells was further increased.
These different patterns of lethal effects in the bystander populations as a function of
dose or activity concentration delivered to the directly-irradiated cells were interpreted as
LET-dependent differences and were, in turn, distinctly different from the saturation
behavior observed with the low-LET
60Co
gamma ray irradiation [96]. The authors
suggested that exposure of the tumor cells to high activity concentrations of the high-LET
radionuclides (1231
21IAt)
could inhibit the cellular ability to generate bystander signals.
This study showed that: (i) the effects were cell line independent, since either of the cell
types showed the same bystander responses; (ii) survival fraction of the bystander cells
was LET dependent; and (iii) the low LET radiation qualities were more effective at
causing cell death in the bystander cells than the high LET particles.
Using the transwell co-culture system, Yang et al. showed similar 2-fold increases
in p21waflexpression, y-H2AX foci, and MN and a 20% survival fraction decrease in
bystander AGO 1522 normal fibroblasts co-cultured with either 250 kVp x-rays or ironparticles (51 keV/Rpm, 1 GeV/amu) irradiated AG01522 cells [43, 63, 64]. Cu-Zn
superoxide dismutase (SOD) and catalase blocked the y-H2AX foci induction in
bystander fibroblasts co-cultured with either iron or x-ray irradiated fibroblasts.
Therefore, ROS were involved in the medium-mediated bystander signal from both
radiation types. This study suggested that the bystander responses in the AGO1522 cells
were LET independent.
1.2.1.8. Bystander Effects in Tumor Cells
Radiation has been used for therapeutic purposes for many years. A concern
during tumor radiation therapy is the possibility of damage to normal cells and tissues.
Bystander effect studies, while initially focused on normal cell effects for radiation
protection purposes, have also been performed on tumor cell lines. Bystander effects in
tumor cells that were co-cultured with directly-irradiated tumor cells have been shown [810, 38, 91].
There have also been a few studies showing increased damage in bystander
normal cells that were co-cultured with directly-irradiated tumor cells. Using the
radioresistant glioblastoma T98G cell line, Shao et al. showed significant bystander
effects in co-cultured normal human fibroblast (AG01522) cells [8, 14]. Dishes were set
up with a population of T98G cells and a population of AGO 1522 cells roughly 5 mm
apart. Using an alpha particle microbeam, 1 or 10 T98 cells were irradiated with 1 or 5
alpha particles through the cytoplasm and MN induction was assayed in the bystander
AG01522 cells. A 2-fold MN induction increase was seen in the AG01522 fibroblasts
and the induction was independent of both the number of T98G cells irradiated and the
number of alpha particle traversals. A 2-fold increase in MN in the AGO 1522 bystander
cells was also seen when either one or 100% of the T98G cells were irradiated in the
nucleus with 1 or 5 alpha particles. Interestingly, only a 1-fold increase was seen when
the T98G cells were the bystanders and one or 100% of the T98G cells were irradiated in
the nucleus with 1 or 5 alpha particles. The addition of filipin, an inhibitor of
glycosphingolipid-enriched membrane microdomains or rafts, abolished all MN
bystander effects in the AG01522 cells. A 30% increase in NO production was detected
in bystander AG01522 cells co-cultured with irradiated tumor cells. In contrast, no
increase in NO production was seen in bystander T98G cells co-cultured with directlyirradiated fibroblasts. Supporting these results, the NO scavenger PTIO blocked the MN
bystander effect in the AG01522 cells co-cultured with the irradiated tumor cells but did
not block the MN bystander effect in the tumor cells that were co-cultured with irradiated
normal fibroblasts. MN bystander effects in tumor cells co-cultured with irradiated
fibroblasts were abolished by the addition of either DMSO or SOD and catalase. These
data suggested that whether the bystander signal was emitted from irradiated tumor cells
or from fibroblasts, the response in the bystander cells was cell-type dependent. Most
importantly, this study suggested that in vitro, irradiated tumor cells were capable of
inducing damage in neighboring normal cells via bystander effects.
Using the medium-mediated methodology, Burdak-Rothkamm et al. also carried
out tumor and normal cell co-culture bystander experiments [16]. T98G or normal human
astrocyte (NHA) cells were grown on 6-well plates and irradiated with a 2 Gy dose of
250 kVp x-rays. Immediately following the irradiation, a coverslip with either T98G or
AG01522 cells was added to the 6-well plates. Following a 90 min co-culture time, the
bystander cells were removed and assayed for y-H2AX foci induction. A significant yH2AX foci induction was seen in bystander NHA cells co-cultured with directlyirradiated T98G cells, and the effect was abolished by DMSO but not by filipin.
Additionally, a significant y-H2AX foci induction was seen in bystander T98G cells cocultured with directly-irradiated NHA cells, and both DMSO and filipin abolished the
effect. The results showed that irradiated tumor cells were capable of inducing significant
double strand breaks in neighboring unirradiated normal cells and that the bystander
effect was ROS dependent.
1.2.1.9.
Temperature Effects on the Bystander Signal
The effect of temperature on cellular responses to direct irradiation has been
widely studied [31-33]. Nearly all studies on the bystander effect have been performed at
room temperature. There is, however, one report on a bystander effect using lowtemperature radiation exposure. In one of the earlier studies from the Mothersill and
Seymour research group, they studied if irradiation at low temperature influenced the
production of the bystander signal. Immortal but nontransformed human keratinocyte
cells (HaCAT) were irradiated with 5 Gy of gamma-rays. One hour after the irradiation,
ICCM was removed from the irradiated cells, filtered, and added to a population of
unirradiated HaCAT cells. The bystander cells were then assayed for survival fraction.
When ICCM was added to the bystander HaCAT cells, a consistent reduction in survival
to -70% was observed. To find out whether the signal in the ICCM was protein-like, the
medium was either heated to 70'C or frozen to -20'C for one hour. The ICCM was then
brought to 37"C and added to the bystander population. Heating of the medium
effectively removed the decrease in survival fraction but cooling of the medium still
induced a bystander effect in the keratinocytes. Since heat denatures proteins, the
bystander-signaling pathway in the directly irradiated HaCAT cells must include a stable
substance that is protein-like. Expanding further on these results, the HaCAT cells were
held on ice at 00C for one hour, irradiated with 5.0 Gy of gamma-rays, held on ice for one
additional hour and then the cold ICCM was removed, warmed to 370C, and placed on
the bystander cells. The directly-irradiated cells exhibited the reduction in survival
fraction to -10% after 5 Gy, regardless of whether irradiated on ice or at room
temperature. The cooling of the cells did not significantly affect the plating efficiencies of
the non-irradiated controls. Bystander cells that received the ICCM that had been held on
ice for 1 hr before, during, and 1 hr after the irradiation did not show a reduction in
survival fraction. The cold temperature prevented the emission of the factor(s) that
induced the survival fraction losses in bystander cells. Biological processes, such as
DNA repair, are slowed down when cell populations are held at cold temperatures [99].
This suggested that the bystander signal, or a component of the signal transduction
pathway, is biological in nature.
1.2.2.
In vivo Bystander effects
For bystander work in vitro to be of any relevance to radiation protection and
therapy applications, the effect must be shown to occur in vivo. Mothersill et al. have
shown a calcium signaling dependent bystander effect in the immortalized human
keratinocyte cells HPV-G in vitro [59, 89, 93]. To expand on their previous results, the
group assessed whether in vivo whole body irradiations of mice led to an in vitro
bystander effect [100]. One hour after adult C57BL/6 mice received a whole body dose
of 0.5 Gy gamma-rays, the bladder was removed, processed to create a single cell
suspension, and plated in a dish. Twenty-four hours later, the ICCM from the bladder
cells was removed, filtered, and added to either unirradiated HPV-G cells or bladder
cells. HPV-G cells that received ICCM from the bladder cells irradiated in vivo exhibited
an increase in intracellular Ca2+ and a decrease in mitochondrial potential, both of which
are markers of the apoptosis pathway. Additionally, both the HPV-G and mouse bladder
bystander cells exhibited a significant increase in apoptosis 72 hours after having
received ICCM from in vivo irradiated bladder cells. While this again was not a true in
vivo study, it illustrated that cells derived from mice that received whole body irradiation,
could emit a bystander signal. Several other research groups have conducted "in vivo
bystander effect" studies with different approaches (described below).
1.2.2.1.
Bystander Effects in 3-D Pelleted Cell Clusters
An experimental approach that has been utilized to imitate a tissue
microenvironment in vivo is the three dimensional multicellular model. Persaud et al.
used this 3-D in vivo cluster model to study bystander effects from low-LET radiation
exposures [101, 102]. Chinese hamster cells (CHO) were radiolabeled with tritiated
thymidine ( [3H]dTTP), a beta emitter with a range of 1 gm in water [101]. The
radiolabeled CHO cells were then mixed at a ratio of 1:4 with unlabeled AL cells, which
are a human-hamster hybrid cell that contain the standard hamster chromosomes and a
single copy of human chromosome 11. The two cell lines were then pelleted to produce a
cluster. The cluster was held for 24 hrs at 11 C to inhibit cell growth but to allow for
accumulation of dose in the CHO cells and for bystander signals to propagate to the AL
cells. It was verified that in these 3D clusters, the cells do communicate between each
other via gap junctions. After 24 hrs, a magnetic separation protocol was performed to
separate the two cell lines. A CD59 antibody, which binds only to the cell surface antigen
on AL cells, was added to the cluster. Magnetic beads coated with rabbit anti-mouse IgG
(secondary antibody to CD59) were also added to the cluster. The cells were passed
through separation columns between magnets, which caused the separation of the two
cell populations. The AL bystander cells were further processed for survival fraction and
for mutation in the CD59 gene, which is present in chromosome 11. A reduction of
survival fraction to 70% as well as a 4-14 fold increase in mutation in the CD59 gene was
detected in the AL bystander cells. Mutations in the CD59 gene in the bystander cells
showed a high frequency of at least three marker losses. When lindane, a gap junction
inhibitor, or DMSO, a general radical scavenger, was added to the clusters during the coculture time, a reduction of 50% of mutation frequency in the CD59 gene was seen in the
bystander AL cells. The bystander effect must then be due to a combination of both gap
junction and medium-mediated signals. This study showed that bystander effects
occurred in a 3-D cluster model. This 3-D cluster model was used to mimic an in vivo
tissue environment. Signaling in vivo in human tissues are much more complex than in
this 3-D cluster model.
1.2.2.2. Bystander Effects in an Artificial 3-D Tissue Model
Multicellular 3-D normal human tissue model have been shown to maintain in
vivo differentiation and metabolic patterns and to form gap junctions. The usage of these
normal three-dimensional human tissues allows for bystander effects studies that are
closer to an in vivo approach the 3-D normal tissues will have a more complex level of
signaling. Belyakov et al. studied the bystander effect in two 3-dimensional normal
human tissue models [103]. The first tissue model is the 75 ýtm thick EPI-200, which is
composed entirely of keratinocyte cells and models the epidermis layer of the skin. The
second model used was the 700 gm full thickness skin EFT-300, which contains an
epidermal layer of only keratinocytes and a dermal layer of fibroblasts. Both the EPI-200
and the EFT-300 are commercially available from MatTek (Ashland, MA). A 7.2 MeV
alpha- particle microbeam was used to irradiate a total of 400-800 EPI-200 keratinocyte
cells in a single thin vertical plane with an average dose of 1 Gy to each cell. In the EFT300 model, the alpha-particle microbeam was used in a similar fashion but either the
keratinocyte cells (top) or the fibroblast population (bottom) was irradiated in a single
thin vertical plane. At 72 hours post irradiation, micronuclei or apoptosis induction was
measured in unirradiated cells at distances, in parallel sections away from the irradiation
plane of increments of 5 jm, up to 1100 jm. In the EPI-200 tissue model, a 3-fold
increase in apoptosis induction was detected in unirradiated keratinocytes up to 1.1 mm
away from the plane of irradiated keratinocytes. Additionally, a 2-fold increase in
micronuclei induction was detected in the unirradiated keratinocytes up to 600 gim away
from the irradiation plane. In the EFT-300 tissue model, a 2-fold increase in apoptosis
induction was seen in the unirradiated keratinocytes up to 1100 gim away from the plane
of irradiated keratinocytes. In contrast, no apoptosis increase was seen in the unirradiated
keratinocytes at any distance away from the plane of irradiated fibroblasts. These data
suggest that the irradiated keratinocytes release a signal that induced a response in the
keratinocytes in the EFT-300 tissue model. Based on the data presented, it was unknown
whether the fibroblasts do not release a bystander signal at all, or whether the
keratinocytes just do not respond to the bystander signal that the fibroblasts do release.
The data did show that long-ranged bystander effects occurred in these 3-D normal tissue
models.
Using a similar microbeam methodology, Sedelnikova et al. further studied the
bystander effects in 3-D normal tissue models [75]. The EFT-300 skin model was used
and as well as the Epi-Airway (Air-100) tissue system. Air-100 is a 40 gpm thick highly
differentiated tissue composed of four layers of normal, human-derived
tracheal/bronchial epithelial cells. Air-100 closely resembles the epithelial tissue of the
respiratory tract. In the EFT-300 skin model, again either the fibroblasts or the
keratinocytes were directly-irradiated. The bottom layer of each tissue model was
irradiated with a 7.0 MeV alpha particle microbeam. The range of these alpha-particles
was approximately 31 pm. Bystander effects were measured and totaled in cells, in
increments of 5-10 gm, up to 1000 pm away using perpendicular cuts to the irradiated
plane of the tissue. These cuts were composed of a section of irradiated cells (bottom)
and bystander cells (top). A schematic of this system is shown in Figure 1-2.
Cut
31 mn
Figure 1-2. Schematic of 3-D normal tissue model irradiations with alpha-particle
microbeams. Arrows represent the beam of alpha-particles. For assaying, a
perpendicular cut to the irradiated plane was performed. Diagram is reproduced from
Sedelnikova et al. [75].
A 3-fold increase in y-H2AX foci induction that persisted up to 7 days post-irradiation
was detected in the bystander cells 1000 glm away from the irradiated plane in the Air100 epithelium. A 3-4 fold increase in y-H2AX foci induction that persisted up to 7 days
post-irradiation was detected in the bystander keratinocyte cells 1000 im away from the
plane of irradiated fibroblasts in the EFT-300 tissues. In vitro studies performed by the
same group had detected y-H2AX foci inductions in monolayers of bystander fibroblasts
that had been co-cultured with directly irradiated fibroblasts for 48 hrs; y-H2AX foci
inductions returned to background levels by 72 hrs [77]. In the tissue models, the levels
of y-H2AX foci inductions remained elevated in the bystander cells until 7 days. The
longer persistence of y-H2AX foci induction in the bystander population could be
attributed to a more complex level of signaling in the 3-D normal tissue system than in
monolayers of cells. A smaller 2-4 fold increase in y-H2AX foci induction that persisted
up to 7 days post-irradiation was detected in the bystander keratinocyte cells 1000 gLm
away from the plane of irradiated keratinocytes in the EFT-300 tissues. In summary, all
three types of bystander cells, in both tissue models, showed a bystander y-H2AX foci
induction that involved roughly 40%-60% of the entire bystander population studied.
Additionally, a 6-fold apoptosis and a 2.5-fold micronuclei increase was seen in the
bystander cells in the Air-100 tissue model that persisted up to 7 days post-irradiation. In
the EFT-300 tissues, both a 2.7-fold apoptosis and 2-fold micronuclei increase persisted
for 7 days in the bystander keratinocytes distal to the irradiated fibroblasts. Unlike the
Belyakov data, there was an induction of a bystander response in the keratinocytes from
the directly-irradiated fibroblasts, suggesting that the fibroblasts indeed emitted a
bystander signal. A possible explanation for this discrepancy was this study had a more
sensitive assay for apoptosis.
Most importantly, the Air-100 tissue microbeam irradiation is a close
experimental representation of a radon type of inhalation exposure to the lungs. These
two studies therefore suggest that significant increases in damage to the epithelial tissue
can be induced due to bystander effects. However, both apoptosis and senescence can be
viewed as protective measures for tumor suppression. The bigger question is whether the
bystander effect is a response to a hazardous environment (i.e. radiation), and selectively
clearing damaged cells in the population to allow for healthy cells to replicate or whether
the bystander effect is stochastically damaging additional cells in the population and can
therefore be viewed as a hazardous effect.
1.2.2.3. Bystander Effects in Mice
The induction of DNA double strand breaks in bystander cells has been widely
shown in vitro, and also in an in vivo study in mice. Koturbash et al. irradiated mice with
a 1 Gy x-ray dose to only the left side body; lead (2.5 mm) was used to shield the right
side of the body [104]. In the half-body irradiations, bystander (shielded side) ventral
skin tissue from at least 7 mm away from the irradiated side was removed and assayed for
genetic and epigenetic changes. The bystander half of the mouse's body would have
received a low 1.3 mGy dose of x-rays due to scatter. Two control groups of mice were
either irradiated with 1.3 mGy as controls for the effects from scatter or were completely
shielded. Six hours after half-body irradiations, a 1.3-fold increase in y-H2AX foci
induction was detected in the bystander skin tissue removed from the unirradiated side of
the mice. By four days, this increase in y-H2AX foci formation was no longer detected in
the bystander tissue. Interestingly, RAD51 expression, an important protein in HR repair,
was significantly increased in the bystander tissue at 4 h and 4 days post-irradiation.
Epigenetic changes such as DNA methylation and histone alterations have been shown to
be involved in cancer development and genomic instability [104]. A slight but not
significant decrease in DNA methylation was observed in the bystander skin tissue 6 h
post irradiation in the half-body mouse irradiation experiment [104]. DNA methylation
patterns are maintained by three DNA methyltransferases, DNMT1, DNMT3a, and
DNMT3b. DNMT1 has been shown to copy the current methylation patterns of DNA
after replication [105]. A 1.4-1.8 fold increase in DNMT1 was observed in the bystander
tissue 6 h and 4 days post irradiation. Additionally, DNMT3a and DNMT3b, which are
involved with new methylation patterns, were both observed to be slightly down
regulated in the bystander tissues. Down regulation of DNMT3a and DNMT3b have been
shown correlate with an early injury response. The methyl Cp-G-binding domain (MBD)
family consisting of MeCP2, MBD1, MBD2, and MBD3, mediate DNA methylation. In
the bystander skin tissue, both MeCP2 and MBD2 showed a 2-fold increase in
upregulation that persisted until 4 days post irradiation. This result complimented the
observation of the DMNT1 induction in the bystander skin tissue. The fully-shielded
mice from a 1 Gy gamma-ray exposure and the mice that received a whole-body 1.3 mGy
x-ray dose did not show any changes in y-H2AX, Rad5 1,DNMT1, MeCP2, or MBD2
levels. To test if the organs that were irradiated influenced the bystander data, a group of
mice had their left side shielded, and instead the right side was irradiated. RAD51 and
DNMT1 were both upregulated to similar levels regardless of which side of the mouse
was irradiated. However MBD2 and McCP2 was only induced in bystander skin when
the left side of the body was irradiated, but not from the right side. One possible
explanation was that when the right side of the mouse was shielded, the spleen was
exposed to gamma-ray exposures. When the left side was shielded, more of the liver was
exposed. These results suggest that the bystander response was dependent on which
organs were directly-irradiated. To summarize the studies findings: (i) increased levels of
double strand breaks were present 6 hours post-irradiation in the bystander skin tissue;
(ii) significant HR repair of the DNA double strand breaks occurred in the bystander
tissue up to 4 days post-irradiation; (iii) DNA methylation genes were involved in the
bystander response; (iv) the bystander response was dependent on which organs were
directly-irradiated; and (v) bystander effects did occur in vivo. The complexity of in vivo
signaling was seen in this study by the tissue/organ differences in bystander responses.
1.2.2.4. LET Dependence of in vivo Bystander Effects
In vitro studies of the LET-dependence of the bystander effect, as discussed above
in section 1.2.1.7, have shown to be inconclusive. There are only a few studies that
looked at the LET dependence of the bystander effect using in vivo experiments. Wright
et al. used a unique bone marrow transplant experiment to show in vivo bystander effects
with two different radiation qualities [106, 107]. Bone marrow (40XY cells), obtained
from CBA/H mice, was irradiated with either a 4 Gy gamma-ray dose or a 0.5 Gy neutron
dose, which contained a 0.125 Gy gamma-ray component. Survival fraction levels of 5%
and 25% were observed in the directly-irradiated 40XY cells after gamma-rays or neutron
exposures, respectively. Immediately following the irradiation, 40XY cells were mixed in
a ratio of 2:3 with unirradiated congenic CBA strain bone marrow (40XYT6T6 cells),
which are homozygous for the stable T6 reciprocal translocation between chromosome
14 and 15. The two small chromosomes in 40XYT6T6 cells allow these cells to be
recognized as the unirradiated bystander cells. The 40XY-40XYT6T6-cell mixture was
then injected into unirradiated recipient mice within one hour after the external beam
irradiation of the 40XY cells. For the gamma-ray irradiation experiments, at days 10, 30
and 100, bone marrow was removed from the mice, and the progeny of the original
bystander 40XYT6T6 cells was assayed for chromosomal aberrations. At all time points
assayed, a 3-fold increase in cytogenetic aberrations was detected in the progeny of the
original bystander 40XYT6T6 cells. In comparison, a 5-fold increase in cytogenetic
aberrations was detected in the progeny of the 40XYT6T6 cells at all time points assayed
between 3-12 months post neutron irradiation. These data suggest that: (i) an in vivo
transmission of a bystander signal occurred between the irradiated 40XY bone marrow
cells and the bystander 40XYT6T6 cells; and that (ii) the high-LET neutron irradiation
was more effective at inducing bystander effects than the low-LET gamma irradiation of
the bone marrow.
One concern with the Wright study is that perhaps the bystander effects observed
were due to a bystander signal being emitted during the mixing time period, and therefore
is an in vitro effect instead of a true in vivo effect. To test for this, the Wright group
performed a control in vitro bystander study. Bone marrow was removed from mice and
irradiated with a gamma-ray dose of 4.0 Gy [106]. ICCM was removed and added to a
bystander population of unirradiated bone marrow cells. Four hours later, the in vitro
bystander bone marrow cells were added to an unirradiated recipient mouse. At 30 d and
100 d post bone marrow transplant, the progeny of the bystander cells were harvested
from the recipient mice and assayed for chromosome aberrations. For this in vitro/in vivo
bystander study, no significant induction of cytogenetic aberrations was detected in the
progeny of the bystander bone marrow cells. These results showed that the in vivo
bystander signals were the cause of the chromosomal aberration bystander effects
detected in the progeny of the original unirradiated bone marrow cells and were not an
artifact of an in vitro effect that occurred prior to injection into the mouse. The in vitro
Wright study contradicts other bystander reports in which cells removed from a mouse
showed significant bystander effects in an in vitro type of bystander experiment [100,
108]. Lorimore et. al. irradiated bone marrow cells obtained from male mice with alpha
particles and showed increased levels of in vitro chromosome aberrations in the bystander
bone marrow samples [108].
Most of the in vivo bystander studies have approached the effect from a radiation
protection point of view. Kassis et al. instead looked for an in vivo bystander effect with
tumor cells [11, 12]. The DNA of human LS 174T colon adenocarcinoma cells was
labeled with lethal levels of either 5-[125I] iodo-2'-deoxyuridine (1 25IUdR) or 5-[1231]
iodo-2'-deoxyuridine (123IUdR). Both iodine isotopes are auger electron emitters with
similar x-ray emission spectra.
60.5 day half-life.
123I
1251
emits 20 low-energy electrons per decay and has a
emits on average 11 low-energy electrons per decay, has a 13.3 h
half-life, and will have a much higher specific activity than 1251. The Auger electrons
from both iodine isotopes have ranges that confine approximately 99% of the energy
deposition within the individual radiolabeled cell and will thus lethally irradiate only the
labeled cell. In a separate experiment, LS 174T cells were irradiated with a lethal gamma-
ray irradiation dose of 5 or 20 Gy. Various mixtures of either gamma-ray irradiated, Ilabeled and unlabeled tumor cells were injected in nude mice to form subcutaneous
tumors. At days 11-15 post injection, the tumor volume was measured. The gamma-ray
irradiated LS174T cells exhibited no in vivo bystander effects. Mice that were injected
with mixtures of 125I-labeled and unlabeled LS 174T cells exhibited a 3-fold decrease in
tumor volume as compared to mice which were injected with only unlabeled tumor cells.
This inhibitory effect on tumor growth was attributed to additional cell death from
bystander effects that had occurred in the tumor. Mice that were injected with mixtures of
123I-labeled
and unlabeled LS 174T cells exhibited a 2-fold increasein tumor volume as
compared to mice that were injected with only unlabeled tumor cells. The bystander
signal from the 123I-labeled cells was involved in a stimulatory way. Stimulatory effects
have also been shown in in vitro bystander studies [60, 82, 92]. Mothersill et al. observed
survival fraction levels of -150-200% in bystander cells, which indicated a significant
increase in cell proliferation in vitro [82].
To explore the in vivo results further, the Kassis group performed in vitro
experiments with both iodine isotopes [12]. Mixtures of labeled and unlabeled tumor
cells were plated into 6-well plates. After 4 days, the medium was removed, centrifuged,
and frozen at -70°C for a week in order to allow for any residual isotopes to decay. The
medium was then added to a new population of unlabeled tumor cells. The bystander
tumor cells that received medium from 125I-labeled cells exhibited a 2.5-fold reduction in
cell growth. The bystander tumor cells that received medium from
123I-
labeled cells
exhibited a 1.8-fold increase in cell growth. The in vitro results confirmed the in vivo
results. Further analysis on the supematants from
125I-labeled
cells detected significant
concentrations of the anti-angiogenic factors TIMP1 and TIMP2. In the 123I-labeled cells,
increased levels of these two factors were not detected, but levels of angiogenin, an
angiogenic factor, were significantly increased in the supernatant. The increase in levels
of these factors supported the growth results observed in the unlabeled bystander cells.
The dose rate in the
123I-labeled
cells was - 109-times greater than in the
125I-labeled
cells.
The number of decays in the 123I-labeled cells was 5.6-fold greater than in the 125I-labeled
cells. While both isotopes have similar Auger electron and x-ray emission spectra, these
data suggest that the dose rate played a role as to the type of factors emitted as a
bystander signal.
1.3 Thesis Work
Most in vitro and in vivo bystander effect studies irradiate normal cells and look
for damage in a separate bystander population of normal cells. There are relatively few
bystander effect studies using tumor cells. However, there is a rationale for such studies.
Clinically, the vast majority of tumors are treated with low-LET radiation, however,
given the increasing applications of high-LET particle beams in radiation therapy [109]
and the continuing development of targeted therapy using high-LET radionuclides [96,
110, 111], investigation of the LET-dependence of bystander effects following tumor
irradiation takes on increased relevance. Also, during tumor therapy, bystander signals
from irradiated tumor cells could produce effects in nearby normal cells or in nearby
tumor cells. This thesis work represents an in vitro model of the type of possible
bystander effects that might occur in vivo during tumor therapy, such as systemic
radioimmunotherapy of micrometastatic prostate cancer using alpha-emitting
radionuclides. DU-145 human prostate carcinoma cells have been shown to cause
medium-mediated bystander effects [9, 10]. Previous results have shown significant
induction of MN formation in bystander DU-145 cells that were co-cultured with alphaparticle irradiated DU-145 cells [10]. Interestingly, the MN bystander effect was only
observed if the co-cultured DU-145 cells were present in the medium during the
irradiation; adding the inserts even 1 min after the irradiation abolished the MN
formation bystander effect. The bystander signal was abolished by DMSO, a general
radical scavenger, but not by PTIO, a NO scavenger. This thesis now includes parallel
irradiations with 250 kVp x-rays to provide a low-LET comparison to the alpha particle
results. Additionally, AGO 1522 normal human fibroblasts were included as a second cocultured bystander cell line for comparison with the bystander effects in the DU-145
prostate tumor cells. Three endpoints were monitored in each of these co-cultured
bystander cell lines: MN formation, y-H2AX foci formation and survival fraction.
Irradiations of the DU-145 cells were also performed on ice to further probe the nature of
the bystander signal. This matrix of experiments allows us to address three questions: (1)
Does irradiation of DU-145 prostate carcinoma cells cause bystander effects in the
AG01522 normal human fibroblasts? (2) Are there LET-dependent differences in
bystander effects between 250 kVp x-rays and alpha particles? and (3) Do irradiations at
low temperatures influence the emission of a bystander signal?
Chapter 2
Materials and Methods
2.1.
Cell Lines
The human diploid skin fibroblast primary cell line, AG01522, was obtained from
the Genetic Cell Respository at the Coriell Institute for Medical Research (Camden, NJ,
USA). AG01522 cells were grown at 370 C in a humidified atmosphere of 95% air and
5% CO 2 using ca-modified MEM (Sigma) supplemented with 20% fetal bovine serum
(Hyclone), 100 g.g/ml streptomycin and 100 U/ml penicillin. AG01522 cells were grown
to confluency prior to usage for experiments and the culture was re-started from frozen
stocks after 12 passages. The human prostate carcinoma (metastatic) cell line, DU-145,
was obtained from the American Type Culture Collection (Manassas, VA). DU-145 cells
were grown at 370 C in a humidified atmosphere of 95% air and 5% CO 2 using Eagle's
minimum essential medium containing Earle's BSS, 2 mM L-glutamine (MEM/EBSS,
Hyclone) supplemented with 1.0 mM sodium pyruvate, 0.1 mM non-essential amino
acids, 1.5 g/liter sodium bicarbonate, and 14% fetal bovine serum (Sigma). All
experiments with DU-145 prostate tumor cells used cells in exponential growth. All
plating for experiments was done 24 hours prior to irradiation. The AGO1522 and DU-
145 cells were harvested by trypsinization, counted and replated at the appropriate
densities for each experiment, as described below.
2.2.
Cell Irradiation
The alpha particle source was described previously [10]. The alpha source is a
sealed planar, custom-manufactured
24 1
Am foil (NRD, LLC, Grand Island, NY).
Powdered americium dioxide was mixed with gold to form a 0.5 jpm thick foil to form the
active layer. The active layer was then positioned in between two 0.75-1.0 jtm thick gold
layers and the
24 1
Am plus gold layers were attached to a -175 jtm silver backing.
Therefore the alpha particles must pass through -1.5 jtm thick layer of gold to exit the
foil. . The alpha particles emerged from the
241Am
foil with an average energy of 3.98
MeV and an average LET of 127 keV/jpm. The activity of the
241Am
source is 370
kBq/cm 2. The dose rate at the cell position on the mylar membrane is 1.2 Gy/min [10].
The cell irradiator comprises the
241Am
foil, a shutter, and a machined collar to position
the stainless steel mylar dish above the shutter. The air gap between the
24 1
Am foil and
the mylar layer is 5 mm. At the cell position on the mylar membrane, the dose rate from
the 60 keV gamma rays emitted during 241Am decay is negligible (-10 - 6 Gy/min) [10].
For alpha particle irradiations, a custom-made cell culture dish with a replaceable
mylar bottom was used. This culture dish has been described previously and is shown
schematically in Figure 2-1. Briefly, a stainless steel cylinder was machined to allow
placement of 1.4 pm-thick mylar across the bottom, creating a 3.81 cm diameter growing
surface at the bottom of the dish. The mylar was held in place by a secondary outer
stainless steel cylinder fitted with a Vinton rubber o-ring between the two cylinders [10,
112]. The mylar dishes were sterilized in an autoclave and covered with standard 60 mm
diameter plastic Petri dish covers during use to ensure continuous sterile conditions. The
mylar membrane was treated with FNC Coating Mix (BRFF AF-10, AthenaES,
Baltimore, MD) to aid in cell adhesion. Figure 2-la shows cells plated directly on the
mylar coated layer. The bystander cells, which were plated on a coverslip, were placed
into an insert holder that was positioned 1 mm above the mylar layer. The bystander cells
were therefore sharing the same medium as the directly-irradiated cells. This setup tested
for medium-mediated bystander effects. Figure 2-1b shows the steel mylar dish
positioned above the alpha foil (not visible).
Directly
irradiated cells
Bystander c
on mylar layer
(1.4 jim)
Imm
5 mm
g
air gap
a particle
source
I
Figure 2-1. Alpha particle co-culture
irradiation system. (a) schematic
of the insert co-culture system for
alpha particle irradiations. (b) the steel
mylar dish on top of the alpha source,
which is not visible
X-ray irradiations were performed using a Phillips RT250 unit, operating at 250
kVp and 12 mA with 0.4 mm Sn plus 0.25 mm Cu filtration and a focus-to-target distance
of 32 cm. The x-ray dose rate in the six well plates was 1.0 ± 0.03 Gy/min.
2.3.
Co-culture Experiments
For alpha particle irradiations, 3-5x10 5 DU-145 cells were plated 24 hrs prior to
irradiation on the coated mylar membrane and allowed to attach overnight. Similarly for
the x-ray irradiations, 1.0-1.3 x 105 DU-145 tumor cells were plated per well in 6-well
plates (Falcon) 24 hrs prior to irradiation and allowed to attach overnight. Tumor cells
plated on a mylar layer or on 6-well plates served as the directly-irradiated cells for alpha
particle and x-ray irradiations respectively. For both radiation quality experiments,
bystander populations of DU-145 or AG01522 cells were prepared on 18 mm-diameter
glass coverslips (VWR International) by plating 1 x 105 cells per coverslip in 12-well
plates (Falcon) and allowing the cells to attach overnight. Bystander cells were present
on the coverslip held in the insert above the mylar layer (see Fig 2-la) or above the 6well plates. On the day of irradiation, the medium was changed in the mylar dish or the 6well plates as well as for the bystander cells growing on coverslips in 12-well plates.
When AG01522 fibroblasts were to be co-cultured with irradiated DU-145 tumor cells,
both cell lines were allowed to attach in their respective medium, but, for the tumor cells,
the medium was changed to AG01522 medium just prior to irradiation. Hence, for coculture experiments with AGO 1522 fibroblasts, DU-145 cells were irradiated in AGO 1522
fibroblast medium. A growth-rate experiment was performed to confirm that the
AGO1522 medium caused no changes in DU-145 cell growth fraction. The bystander
cells were then co-cultured with the directly-irradiated cells for 4 hrs (except as noted
below), removed, and then processed for the specific endpoints described below.
The geometry of the two-irradiation approaches produced a limitation and an
opportunity. For x-ray irradiations, the co-cultured bystander cells could not be present
in the medium during the irradiation, and were thus always added within 5 minutes after
the irradiation. During alpha particle irradiations, the range of the alpha particles is
roughly 30 jpm in water. The bystander cells could then be present in the same medium
above the irradiated cells, but still receive no direct dose. Alternatively, the bystander
cells on the inserts could be added after the irradiations (within 5 minutes, or as short as 1
minute) to parallel the x-ray conditions. This provided the opportunity to investigate the
nature of the signal released from alpha particle-irradiated DU-145 cells, i.e., the lifetime
of the medium-mediated signal, by comparing the results when the bystander cells were
added before or just after the irradiation [10].
Thus, in all x-ray experiments, bystander cells and directly-irradiated cells were
combined for co-culturing - 5 min after irradiation. In alpha particle experiments, the
bystander cells could be added to the medium above the irradiated cells either before or
after irradiation: the experimental conditions used are specified in the results sections
below.
2.4.
Radical Scavenger Experiments
Radical scavenger experiments used either the nitric oxide scavenger 2-phenyl-
4,4,5,5-tetramethylimidazoline-1-oxyl 3-oxide (PTIO: MP Biomedicals, Inc) at a
concentration of 20 jM or dimethyl sulfoxide (DMSO: Sigma) at 0.5% vol/vol. The
scavengers were added 5 min prior to irradiation and were present in the medium during
the entire co-culture period.
DMSO is non-enzymatic free radical detoxifier and is considered to be the most
powerful scavenger of hydroxyl radicals ('OH). The reaction of DMSO with OH
produces CH 4 and methane sulfinic acid [113, 114]. This reaction is shown in Figure 2-2.
PTIO is a stable radical that scavenges nitric oxide (NO) [115]. The reaction is shown in
Figure 2-3.
O
CH
S=O +-OH -
CH
+ CH 3 -S
CH
OH
Figure 2-2. Chemical reaction of dimethyl sulfoxide (DMSO) with OH'.
0
ii
I
i
R + NO --'-
i
R + NO2
O
I "
I O
0
o
,.............................................................................................................................................
Figure 2-3. 2-phenyl-4,4,5,5-tetramethylimidazoline- 1-oxyl 3-oxide (PTIO) reaction
with NO.
2.5
Bystander experiments with 4oC Cell Irradiations
For the temperature-controlled experiments, cells were plated using the same
method described in Section 2.3 for the micronuclei assay or Section 2.9 for survival
fraction experiments (transwell insert method). On the day of irradiation, the medium for
the DU-145 cells was replaced by 4oC fresh medium. The medium for the bystander cells
was exchanged for 37'C warmed fresh medium. The DU-145 cells were then placed in a
refrigerator and held at 0-4°C for one hour. For x-ray irradiations, the 6-well plates were
then placed and irradiated on ice to ensure that the cells remained at 4'C. Alpha particle
irradiations were done in the refrigerator. After the irradiation, the addition of the
bystander cells was delayed to allow for the directly-irradiated DU-145 cells to warm up
to 25"C in the incubator. The delay was 7 min for x-rays and 15 min for alpha particle
irradiation. This was done to ensure that the bystander cells were never in cold medium at
any time during the entire experiment. After the delay, the bystander cells were added.
The bystander cells were co-cultured with the directly-irradiated DU-145 cells for 4 hrs
for the MN assay or 24 hrs for the survival fraction assay.
For each 4°C temperature-controlled bystander experiment, a 37"C bystander
experiment was done in parallel. On the day of the irradiation, 37'C fresh medium was
added to both the bystander cells and the DU-145 cells. Immediately after the radiation
exposure, the directly-irradiated DU-145 cells were placed in the incubator and held at
37*C for 7 min for x-rays or 15 min for alpha particles. After the delay, the bystander
cells were added and co-cultured with the directly-irradiated DU-145 cells for 4 hrs for
the MN assay or 24 hrs for the survival fraction assay.
2.6
Micronucleus assay
The induction of micronucleus formation in the bystander cells was measured
after irradiation of the DU-145 cells with either x-rays or alpha particles. A chromosome
fragment that is not contained with the rest of the nucleus at cell division forms a
micronucleus. Micronucleus (MN) formation in the bystander cells was measured using
the cytokinesis-block technique [116]. Cytochalasin B is used to block cells in
cytokinesis by hindering the formation of contractile microfilaments and therefore leaves
the cells in a binucleated state. After the directly-irradiated cells and the bystander cells
had been co-cultured for 4 hrs, the coverslips with the bystander cells were removed from
the mylar dish (alpha particles) or the 6-well plates (x-rays) and placed in individual
wells of a 12-well plate (Falcon). The bystander cells received two ml of fresh medium,
respective to each cell line, containing cytochalasin B (Sigma) at a final concentration of
1.5 gg/ml for AG01522 cells and 3.0 gg/ml for DU-145 cells. After 48 hrs for the DU145 cells [10] and 72 hrs for AG01522 fibroblasts [43], the cells were fixed in
methanol:acetic acid (3:1 v/v). After allowing the samples to dry, the cells were stained
with the nuclear stain 4', 6'-diamidimo-2-phenylindole (DAPI: Sigma) at a concentration
of 10 jtg/ml in water. The stained bystander cells, still attached to the coverslips, were
treated with FluoroGuardTM Antifade reagent (Bio-rad); the coverslips were then inverted
and applied to microscope slides. Micronuclei present in binucleated cells were scored
using a fluorescence microscope. At least 500 binucleated cells were scored, using at
least 10 fields of view, from each coverslip.
2.7
y-H2AX Assay
The induction of y-H2AX foci was measured in bystander cells. The histone
H2AX, when phosphorylated at serine 139, is known as y-H2AX and has been shown to
occur at locations of DNA double strand breaks [79]. Induction of y-H2AX in the nucleus
of cells can be detected by immunofluorescence staining that visualizes individual foci of
diameter -0.5-1 jim [117]. The number of cells that exhibit y-H2AX foci can then be
tallied and serve as a measure of double strand break formation in the cell.
Immunofluorescent staining of y-H2AX foci was carried out as previously described by
Yang et al., [43]. At the end of the co-culture period, the bystander cells were rinsed with
PBS, then fixed in 3% (v/v) paraformaldehyde in PBS for 30 min at 40 C. After a 5 min
rinse with 50 mM NH4 Cl and two rinses with PBS, the cells were then permeabilized for
15 min in ice-cold Triton X-100 buffer (50 mM NaCl, 3 mM MgCl 2 , 200 mM sucrose, 10
mM HEPES, pH 7.4, and 0.5% Triton-100). Cells were incubated with 10% goat serum
for 1 hr at 37oC, then incubated for 1 hr at room temperature with antibodies against
phosphorylated histone y-H2AX (Trevigen) using a 1:100 dilution of the antibody in PBS
supplemented with 3% goat serum and 0.1% Triton X-100. After three 10 min washes
with PBS, cells were incubated in 1% BSA for 1 hr at 370 C. After blocking, the cells
were stained with Alexa Flour® 488 goat anti-rabbit IgG secondary antibody (Molecular
Probes) for 45 min at room temperature. Cells were washed twice with PBS, stained with
DAPI (Sigma) for 2 min, followed by two more washes with PBS. The cells were treated
with FluoroGuardTM Antifade reagent (Bio-rad) to preserve the fluorescence stain. A cell
was counted positive for y- H2AX foci induction if it exhibited more than 5 foci in the
nucleus. At least 500 cells were scored from at least 10 fields of view per sample.
2.8
Cell Cycle Assay
It has been shown that y-H2AX foci formation can also occur in S-phase cells at
stalled replication forks [118]. In order to verify that the bystander cells entering into Sphase did not confound the y-H2AX foci formation detected, cell cycle analysis was
carried out using flow cytometry. After the directly-irradiated cells and the bystander
cells had been co-cultured for 4 hrs, the bystander cells were removed from the 6-well
plates and placed in individual wells of a 12-well plate (Falcon). The bystander cells
were then trypsinized, collected into a 15 ml test tube, and washed twice with PBS by
spinning the cells in a centrifuge at 1000 rpm for 4 min. Cells were then vortexed to
maintain a single cell suspension and fixed drop by drop with 100% ice-cold ethanol, and
fixed in 4°C for at least 24 hrs. Cells were then washed in PBS + 1% BSA. Cells were
then stained with 500 ug/ml Propidium Iodide (PI) (Sigma) in 3.8 x 10-2 M sodium
citrate, pH 7.0, 10 mg/ml RNase (Sigma) in 10mM Tris-HC1, pH 7.5, in PBS + 1% BSA
for a total working volume of 1 ml. The cells were then incubated at 37°C for 30 min.
The cells were then analyzed for DNA content using the flow cytometer FACScalibur
(Becton Dickson, Franklin Lake, NJ USA) with the software CellQuest V5.1 Mac OS X.
2.9
Colony Formation Assay
A colony formation assay was used to quantify the surviving fraction of bystander
cells following co-culture with directly-irradiated DU-145 tumor cells. For either alpha
particle or x-ray experiments, the bystander cells (either AGO1522 or DU-145: lx 105
cells per coverslip) were added to the medium above the irradiated cells either just prior
to irradiation or shortly after irradiation depending on the experiment, and co-cultured for
4 hrs. After the 4-hour co-culture, the bystander cells were removed, trypsinized, and
plated in 100 mm Petri dishes (Falcon) at a density of 300 cells per dish. Cells were
incubated for 9 additional days.
For bystander scavenger experiments with x-ray irradiations, the transwell coculture system was used. DU-145 cells (1x105) were plated in 1.0 gm porous membrane
inserts (Falcon) to allow passage of molecules [43]. Bystander fibroblast cells were
plated in the 6-well companion plates (Falcon) at 100 cells/well. After allowing the cells
to attach for 24 hrs, the medium was replaced in both the wells and the companion
inserts. For experiments with scavengers, either of the two radical scavengers was added
to the inserts prior to irradiation and the inserts holding the DU-145 cells were then
irradiated with x-rays. Within 5 min after irradiation, the inserts were added to the
companion wells holding the bystander cells and the cells were co-cultured for 24 hrs.
After 24 hrs, the inserts were removed and discarded, and the bystander cells in the wells
were incubated for 8 additional days.
For all survival experiment, colonies were fixed with methanol and stained with
methylene blue. Colonies containing greater than 50 cells were scored per well or per
dish and the surviving fraction was calculated.
2.10
Statistical Analysis
All data presented in the results section were blindly scored by one observer and
were obtained from at least three independent experiments, unless otherwise stated. The
results of MN and y-H2AX experiments are expressed as the mean of all replicate
measurements and are plotted as the mean ± 1 SD. Colony formation survival fraction
experiments involve the mean of 3-5 replicate dishes or wells per experiment: results are
expressed as the mean ± 1 standard deviation (SD) of all replicate dishes or wells for all
experiments. Student's t-test was used to evaluate differences in the effects measured in
the bystander cells from the non-irradiated control conditions and the bystander cells cocultured with directly-irradiated cells. In scavenger experiments, statistical significance
was judged relative to the same absorbed dose delivered to the directly-targeted cells but
without the scavenger present.
2.11 Experimental Approach Summary
For all experiments, DU-145 prostate carcinoma cells were directly-irradiated with
either x-rays or alpha particles. For each of these two radiation qualities, two co-cultured
bystander cell lines were used (DU-145 human prostate carcinoma and AG01522 normal
human fibroblast). The bystander cells were added to the medium within 5 minutes for xrays, and either before or within 5 min after irradiation for alpha particles. For each of the
4 permutations of radiation quality and bystander cell line, three experimental endpoints
were quantified (y-H2AX foci, MN formation, and survival fraction). This yielded a
matrix of 12 experiments, which is shown in Table 2-1. For each of the 12 experimental
conditions, when a positive bystander effect (significantly greater than control levels) was
detected, the experiments were independently repeated with a radical scavenger present.
Table 2-1. Experimental approach.
Chapter 3
Results
3.1 Micronuclei Formation
Cytochalasin B arrests cells in a binucleated state by permanently blocking them
at the G2/M cell cycle phase. Cells are generally incubated with cytochalasin B for
approximately 1-2 cell cycle times in order to gather the majority of the cells at the
binucleated state. When using cytochalasin B, two parameters must be optimized: the
concentration of cytochalasin B that would be least toxic to the cells and the time of
incubation that would give the maximum levels of binucleated cells. Cytotoxicity
optimization of cytochalasin B for the DU-145 cells has been previously reported [10]. It
was determined that a cytochalasin B concentration of 3.0 jlg/ml was least toxic to the
tumor cells and 48 hrs of incubation time caused the greatest percentage of binucleated
cells. For the AG01522 cells, it has been reported that an incubation time of 72 hrs with a
cytochalasin B concentration of 1.5 gpg/ml was optimal [43]; these conditions were used
in this thesis.
Binucleated cells with or without micronuclei (MN) were visually scored in the
DU-145 or AGO 1522 bystander cells. Figure 3-1 shows a typical immunofluorescence
DAPI-stained field showing binucleated nuclei with MN (arrows showing location)
induction in DU-145 cells directly-irradiated with 2 Gy of x-rays. DAPI is a nuclear stain
so the cytoplasm of the cells is not visible in Figure 3-1 (or in Figure 3-2). In the lower
left region of Figure 3-1, another significant DNA damage marker, an apoptosis bridge
(dashed arrow), occurred between two daughter cells.
The incidence of MN formation in bystander AG01522 cells was measured after 4
hrs co-culture with directly irradiated tumor cells. Figure 3-2 shows MN induction in
bystander AG01522 cells co-cultured with unirradiated (Fig. 3-2A) or 2.0 Gy x-ray
irradiated DU-145 cells (Fig. 3-2B).
Figure 3-1. In situ immunofluorescence detection of micronuclei in 2 Gy x-ray
irradiated DU-145. Single arrows: locations of MN, dashed arrow: apoptosis bridge.
Figure 3-2. In situ flourescene detection of micronuclei in bystander AG01522 cells
co-cultured with DU-145 cells. (A) co-cultured with unirradiated tumor cells; (B) cocultured with 1.0 Gy x-ray irradiated tumor cells. Arrows indicate location of MN.
3.1.1 Micronucleus Formation in Bystander DU-145 Cells
The incidence of MN in the bystander DU-145 tumor cells co-cultured for 4 hrs
with x-ray-irradiated DU-145 tumor cells is shown in Figure 3-3A. The fraction of
binucleated cells containing MN increased as a function of increasing x-ray dose to the
directly-irradiated DU-145 cells, reaching a plateau of 1.8-fold greater than the control
level at 2 Gy (P < 0.01). The presence of either DMSO or PTIO during the irradiation and
the co-culture period reduced the bystander effect to control levels (Fig. 3-3A).
An increased incidence of MN formation in bystander DU-145 tumor cells cocultured with alpha particle-irradiated DU-145 tumor cells has been previously reported
[10]. Furthermore, the bystander signal appeared to be short-lived; the bystander effect
was only observed if the. insert containing the bystander cells was present in the medium
above the directly-irradiated cells during the irradiation [10]. These results were
independently confirmed in the current study (Fig. 3-3B). The increased incidence of
MN in the bystander DU-145 cells plateaued at a level 2-fold greater than the controls at
6 Gy (P < 0.01). The bystander effect was only observed when the insert was present
during the irradiation, not when the insert was added within 5 min after the irradiation,
indicating that a factor(s) in the bystander signal cascade pathway has/have a very short
effective lifetime.
The data shown in Figure 3-3 suggests that there is an LET-dependent difference
in the signal released from x-ray- and alpha particle-irradiated DU-145 prostate
carcinoma cells. In the x-ray experiment (Fig. 3-3A), the bystander DU-145 cells were
added within 5 min after the irradiation: a significant MN-formation bystander effect was
observed. Both DMSO and PTIO reduced the MN induction to control levels, and hence
removed the signal. When alpha particles were used (Fig. 3-3B), the MN-formation
bystander effect was only observed when the bystander DU-145 cells were added before
the irradiation. No bystander effect was seen when the bystander DU-145 cells were
added within 5 min after the irradiation.
~
SX-ray
X-ray+DMSO
-
*
X-rnv+PTIO
J
·-
TT
0
Z
n
v L_.
2.0
1.0
0.06
X-ray dose (Gy)
Cu
r-
SBefore
- After
U,
:::
4E
2E
nu
L
0.1
0.6
1.2
6.0
Alpha particle dose (Gy)
Figure 3-3. Fraction (%) of bystander DU-145 cells containing micronuclei following
4 hr co-culture with DU-145 cells that had been irradiated with x-rays (3-3A) or
alpha particles (3-3B). The x-ray experiments (3A) were carried out in the presence or
absence of DMSO and PTIO scavengers. The alpha particle experiments (3B) were
carried out with the bystander DU-145 cells present during the irradiation (Before) or
added within 5 min after the irradiation (After). Results are the average of at least three
independent experiments. Error bars represent ± 1 SD (** P < 0.01, * P < 0.05 relative
to the 0 dose control).
3.1.2 Micronucleus Formation in Bystander AG01522 Cells
An increased incidence of MN formation was observed in bystander AGO 1522
cells co-cultured for 4 hrs with x-ray-irradiated DU-145 cells; this incidence appeared to
saturate above 1 Gy (Fig. 3-4A). A 1.5-fold bystander response (P < 0.05) was measured
when an x-ray absorbed dose of 2 Gy was delivered to the directly-irradiated DU-145
cells (Fig. 3-4A). DMSO appeared to reduce this bystander effect but the differences
from the corresponding no-scavenger-added data points did not reach statistical
significance (P > 0.05). PTIO blocked the MN bystander effect, but only at the 2 Gy
dose point was the effect statistically significant (P < 0.01).
A 1.8-fold increase (P < 0.05) in the fraction of binucleated cells with MN was
seen in bystander AG01522 fibroblasts that were co-cultured for 4 hrs with alpha
particle-irradiated DU-145 tumor cells (Fig. 3-4B). This bystander effect showed
saturation at doses of 0.6 Gy and above delivered to the directly-irradiated DU-145 cells.
Both PTIO and DMSO reduced the bystander response to control levels. The MN
bystander effect in the AG01522 fibroblasts was observed regardless of whether the
bystander cells on the coverslip were added prior to or after the alpha particle irradiation
of the tumor cells, indicating that the signal is relatively long-lived. The data shown in
Figure 3-4 are for addition of the inserts containing the bystander cells within 5 min after
the irradiations.
SX-rays
SX-rays+DMSO
SX-rays+PTIO
z
0
0.06
2.0
1.0
X-ray dose (Gy)
C4
.O
t_
"O
Salpha
Salpha+DMSO
alpha+PTIO
0
CrJ
CVJ
t
,=:
1.2
6.0
Alpha particle dose (Gy)
Figure 3-4. Fraction (%) of bystander AG01522 cells containing micronuclei
following 4 hr co-culture with DU-145 cells that had been irradiated with x-rays (34A) or alpha particles (3-4B) in the presence or absence of DMSO and PTIO
scavengers. Results are the average of at least three independent experiments. Error
bars represent ± 1 SD. (** P < 0.01, * P < 0.05 relative to the 0 dose control)
3.1.3 Temperature Effects on the Production of a Bystander Signal
Enzymatic processes are severely slowed down when cells are held at lowtemperatures. Cells that are irradiated and held at low temperatures do exhibit slower
DNA repair rates [99]. Low-temperature radiation exposures of cells, however, have been
shown to produce the same level of physical damage as that seen in cells irradiated at
room temperature [35-37, 99]. Hypothermic irradiation conditions could affect the ability
of an irradiated cell to emit a bystander signal. By irradiating cells at low temperatures,
the emission of the bystander signal can be further probed in terms of its chemical or
biological nature.
The MN induction in bystander AG01522 cells was measured as a function of
increased dose delivered to the tumor cells that were directly-irradiated at 40 C. The MN
bystander effect in AG01522 cells plateaued at 1.2 Gy in the 25oC co-cultured
experiments (Figure 3-4A), therefore, this dose was chosen for the 4oC experiments.
In
the 4oC temperature-controlled experiments, the addition of the bystander AGO 1522 cells
was delayed by an additional 15 min to allow for the directly-irradiated cells and medium
to warm to 25oC. This was to ensure that the bystander cells were never added to cold
medium. Irradiation of the DU-145 cells at 4oC significantly (P<0.05) abolished MN
induction in bystander AGO1522 cells (Figure 3-5). This indicates that the low
temperature (4oC) irradiation of the DU-145 cells with alpha particles abolished the
production of the MN bystander signal.
The addition of the DU-145 bystander cells with 250 C directly irradiated DU-145
cells was delayed by 15 min in order to mimic the 4oC temperature-controlled
experiments. Figure 3-5 shows that the 15 min delay in the addition of AGO1522
bystander cells to 25oC alpha particle (1.2 Gy) irradiated DU-145 cells did not
significantly change the MN response as compared to the not delayed bystander samples.
The 25oC directly-irradiated tumor cells were either still emitting a bystander signal or
the already-emitted signal was stable in the medium during this additional delay of the
addition of the bystander cells. Therefore, the bystander signal had not already decayed
during the delay time for the 4oC or 25oC -controlled experiments.
I~.Uul
14.U
-
* 40C
r
0
* 25 C
10.00
K
z
8.00
Ci 6.00
S4.00-
2.00-
0.00
7
0
1.2
Dose (Gy)
Figure 3-5. Fraction (%) of bystander AG01522 cells containing micronuclei
following 4 hr co-culture with DU-145 cells that had been irradiated with alpha
particles at 4°C, 25°C with immediate co-culture initiation or 250C with a 15 min
delay in co-culture initiation. Data points are the mean + 1 SD obtained from at least
three independent experiments. (* P < 0.05 for 4oC effect relative to the 25oC effect with
alpha particles at the same dose)
In the 4oC temperature-controlled experiments using x-rays, the addition of the
bystander AG01522 or DU-145 cells was delayed by an additional 7 min. Preliminary
data shows that x-ray-irradiation of the DU-145 cells at 4oC appears to abolish the MN
induction in bystander AGO 1522 cells (Figure 3-6). The data represents only two
independent experiments. X-ray-irradiation of the DU-145 cells at 40 C did not appear to
abolish the MN induction in bystander DU-145 cells (Figure 3-7).
O.Uv
-
a40C
0 250C
7.00 -6.00
O 25 + delay
-
T
z
I
. 5.00 i
:
" 4.00 '1C
3.00
-
i 2.00
-
0
1.000.00
I
Dose (Gy)
Figure 3-6. Fraction (%) of bystander AG01522 cells containing micronuclei
following 4 hr co-culture with DU-145 cells that had been irradiated with x-rays at
4°C, 25°C with immediate co-culture initiation or 25°C with a 7 min delay in coculture initiation. Data points are the averages obtained from two independent
experiments.
14.00 * 400
12.00 -
*
T
250
250C
C + delay
A 10.00 -
z
8.00 -
S6.00
-
4.00 2.00
0.00 0.00-
1
2
0
Dose (Gy)
Figure 3-7. Fraction (%) of bystander DU-145 cells containing micronuclei following
4 hr co-culture with DU-145 cells that had been irradiated with x-rays at 4°C, 25TC
with immediate co-culture initiation or 25TC with a 7 min delay in co-culture
initiation. Data points are the mean ± 1 SD obtained from at least three independent
experiments.
3.2 y-H2AX Foci Expression
Expression of y-H2AX foci formation, a marker for DNA double strand breaks,
was determined in both directly-irradiated and bystander DU-145 and AGO 1522 cells. yH2AX foci induction was scored visually using a fluorescent antibody. A positive count
for y-H2AX induction was if the cell exhibited greater than 5 individual foci. Figure 3-8
shows y-H2AX foci induction in control DU-145 cells (Fig. 3-8A) and in 5.0 Gy x-ray
irradiated DU-145 cells (Fig. 3-8B). Nonspecific binding of the antibody causes a slight
halo around the nuclei.
Figure 3-8. In situ immunofluorescence detection of y-H2AX foci induction in DU145 cells 1.5 h after x-ray irradiation. (A) unirradiated cells; (B) cells directlyirradiated with 5.0 Gy of x-rays.
3.2.1 Time-course of y-H2AX Induction for Bystander DU-145 Cells
Preliminary experiments were carried out with bystander DU-145 bystander cells
to determine the co-culture time for measurement of y-H2AX foci induction. The
background level of DU- 145 cells with > 5 y-H2AX foci was 1.2 ± 0.7 % (no dose to the
directly-irradiated cells, 2 hrs co-culture). The background level was the average from 8
independent y-H2AX induction experiments in DU-145 cells. This average background
level was used in all of the y-H2AX foci data presented for the DU-145 cells in this
section. Figure 3-9 shows the y-H2AX foci expression time-course for bystander DU-145
cells co-cultured with DU-145 cells that had been irradiated with either 2.0 Gy or 10.0
Gy of x-rays. Only one experiment was performed in the DU-145 bystander cells cocultured with 2.0 Gy or 10.0 Gy x-ray-irradiated DU-145 cells, except for the 4 h time
point (10.0 Gy) at which 3 independent experiments were performed. Bystander DU-145
cells co-cultured with 2.0 Gy x-ray-irradiated DU-145 cells for time periods of 5 min, 30
min, 1, 2 and 4 hrs showed no significant increase in y-H2AX foci induction. Bystander
DU-145 cells co-cultured with 10.0 Gy x-ray-irradiated DU-145 cells for time periods of
5 min, 30 min, 1,2 and 4 hrs showed no significant increase in y-H2AX foci induction.
For the x-ray irradiations, the DU-145 bystander cells were added within 5 min after
exposure. Similarly, Figure 3-10 shows the y-H2AX foci expression time-course for
bystander DU-145 cells co-cultured with DU-145 cells that had been irradiated with 1.2
Gy of alpha particles. The data shown were from one experiment, except for the 1.5 h
time point at which 3 independent experiments were performed. Bystander DU-145 cells
co-cultured with 1.2 Gy alpha-particle-irradiated DU-145 cells showed no significant
increase in y-H2AX foci induction at any of the time points studied. For the alpha
particle irradiations, the DU-145 bystander cells were present during the irradiation, i.e,
the co-culture began just prior to the irradiation. The low induction levels of y-H2AX
foci in the bystander DU-145 cells co-cultured with either x-ray- or alpha-particleirradiated cells suggested that the cells do not readily assemble y-H2AX foci. The data
suggests that y-H2AX is not a good endpoint for bystander studies in the DU-145 cells.
CI
3
m2 Gy
S10 Gy
2.5
2
1.5
10.5
-
0-
r _
I_
0 Dose
0.5
0.083
1
2
Time (Hr)
2 hour
Figure 3-9. Fraction (%) of bystander DU-145 cells containing > 5 y-H2AX foci as a
function of co-culture time with DU-145 cells that had been irradiated with either 2
Gy or 10 Gy of x-rays. Data points presented are from a single experiment, except the 0
dose, from 8 independent experiments, and 4 h (10Gy), from 3 independent experiments.
3.00 2.50 2.00 1.50
1.00
0.50
0.00
--
0 Dose
2 hour
I
!
-
0.17
I
1
1.5
4
6
Time (Hr)
Figure 3-10. Fraction (%) of bystander DU-145 cells containing > 5 y-H2AX foci as
a function of co-culture time with DU-145 cells that had been irradiated with 1.2 Gy
of alpha particles. Data points presented are from a single experiment, except the 0 dose,
from 8 independent experiments, and 1.5 h (1.2Gy), from 3 independent experiments.
100
Since the bystander DU-145 cells showed low induction levels of y-H2AX foci under
the studied conditions, experiments were performed to test y-H2AX foci induction levels
in directly-irradiated DU-145 cells. In Figure 3-11, a 2-fold increase in y-H2AX foci
induction at 4 h post irradiation was seen in DU-145 tumor cells directly-irradiated with
2.0 Gy of x-rays. DU-145 tumor cells irradiated with 10.0 Gy of x-rays showed a 10-fold
increase in y-H2AX foci induction at 2 h post irradiation. In Figure 3-12, at 4 h post
irradiation with 6.0 Gy of alpha-particles, DU-145 cells showed a 10-fold increase in yH2AX foci induction. All results presented in the irradiated DU-145 cells were obtained
from one experiment. These data suggest, along with another published study [119], that
the DU-145 cells induce high induction levels of y-H2AX foci only after direct high dose
exposures.
'U
=
18
S16
S14
12
10
8
6
4
2
0
0
~1~
0 Dose
0.083
0.5
1
2
4
6
Time (Hr)
Figure 3-11. Fraction (%) of DU-145 cells containing > 5 y-H2AX foci as a function
of time after direct irradiation with either 2.0 or 10.0 of x-rays. The data in the
irradiated DU-145 cells were from one experiment.
101
,am% ~
.18
16
1412 S 10861
4
S2
;0- S 0-
;0
T
I
I
Dose
0.083
"r
I
0.5
Time (Hr)
"1"
I
---- T--
2
4
Figure 3-12. Fraction (%) of DU-145 cells containing > 5 y-H2AX foci as a function
of time after direct irradiation with 6.0 Gy of alpha particles. The data in the
irradiated DU-145 cells were from one experiment.
3.2.2 Time-course of y-H2AX Induction in Bystander AG01522 Cells
Preliminary experiments were carried out with AG01522 bystander cells to
determine the co-culture time for measurement of y-H2AX foci induction. Bystander
AG01522 cells were co-cultured with x-ray-irradiated (2.0 Gy) DU-145 cells for time
periods of 5 min, 30 min, 1, 2, and 4 hrs, after which the y-H2AX foci were scored.
Table 3-1 lists the results. The background levels (0 dose - 2 hr co-culture time) were
4.55 and 6.9% of the AGO 1522 cells with greater than 5 y-H2AX foci for these two
experiments. From the two preliminary time course experiments, 4 hrs was the consistent
time that exhibited a 2-fold increase in y-H2AX induction in the bystander AGO 1522
cells. Therefore, 4 hrs was chosen as the co-culture time for the subsequent dose response
study.
102
Deriment 1
# of cells
Coculture
Time (Hr)
0 dose -
Experiment 2
# of cells
with
% of cells
with
y-H2AX
y-H2AX
# of cells
# of cells
with
y-H2AX
% of cells
with
y-H2AX
510
23
4.5
521
36
6.9
512
512
503
520
502
46
28
35
41
43
8.9
5.5
6.9
7.9
8.6
504
514
521
529
514
37
46
49
48
66
7.3
8.6
9.4
9.1
12.8
2 hrs
0.083
0.5
1
2
4
Table 3-1. Fraction (%) of bystander AG01522 cells containing > 5 y-H2AX foci at 5
min-4 hr after co-culture with DU-145 cells that had been irradiated with 2.0 Gy of
x-rays. Data from two experiments. The control (0 dose) sample was processed after 2
hrs of co-culture.
Bystander AGO1522 cells were co-cultured with alpha particle-irradiated (1.2 Gy)
DU-145 cells for time periods of 1, 2, 4, 6 and 8 hrs, after which the y-H2AX foci were
scored. The fraction of bystander AG01522 fibroblasts showing an increase in the
incidence of y-H2AX foci as a function of co-culture time with alpha particle-irradiated
DU-145 cells is shown in Figure 3-13. There was no significant increase in the fraction
of bystander cells with > 5 foci per cell at any co-culture time: the control group showed
1.4%, the values at co-culture times of 1, 2, 4, 6 and 8 hrs were 1.1, 0.9, 1.0, 1.6, and 0.9
% respectively. A co-culture time of 8 hrs was used for the subsequent dose-response
study. For all of the alpha particle irradiations, the bystander AG01522 cells were added
within 5 min after the direct irradiation of the DU-145 cells.
103
I•Pt
2L.5
T
2.00
T
II
1.50
T
-
1.00
T
II
II
S0.50
0.00 0 Dose
2 hr
I
I
II
2
4
I
I
Co-culture Time (Hr)
Figure 3-13. Fraction (%) of bystander AG01522 cells containing > 5 y-H2AX foci at
1-8 hr after co-culture with DU-145 cells that had been irradiated with 1.2 Gy of
alpha particles. Results are the average of at least three independent experiments. Error
bars represent + 1 SD.
3.2.3 y-H2AX Foci Induction in Bystander AG01522 Fibroblasts
The fraction of bystander AGO1522 fibroblasts showing an increase in the
incidence of y-H2AX foci as a function of either direct x-ray or alpha particle dose
delivered to the DU-145 cells is shown in Figure 3-14. The magnitude of the y-H2AX
incidence in the bystander cells in the medium above x-ray-irradiated DU-145 cells
increased linearly up to a level that was 2-fold greater than the control (Fig. 3-14A).
When either of the free radical scavengers DMSO or PTIO was present in the medium
during the irradiation of the DU-145 cells with x-rays, and during the 4-hr postirradiation co-culture period with the bystander AG01522 cells, the bystander effect was
completely blocked at all x-ray doses (Fig. 3-14A).
104
The fraction of bystander AG01522 fibroblasts showing an increase in the
incidence of y-H2AX foci as a function of the direct alpha particle dose delivered to the
DU-145 cells is shown in Figure 3-14B. The bystander AG01522 cells were co-cultured
with the irradiated DU-145 cells for 8 hrs. No increase in y-H2AX foci formation was
observed in the bystander AG01522 fibroblasts at any alpha-particle dose. For all of the
alpha particle irradiations, the bystander AG01522 cells were added within 5 min after
the direct irradiation of the DU-145 cells.
The data presented in Figure 3-14 represent a clear example of an LET-dependent
bystander effect: x-ray irradiation of DU-145 tumor cells produced a significant y-H2AX
bystander effect in AG01522 fibroblasts, alpha particle irradiation did not. In both cases,
co-culture of the bystander AG01522 cells with the directly-irradiated DU-145 cells
began within 5 min after irradiation.
105
SX-rays
7
X-rays+DMSO
X-rays+PTIO
6
5
E
T
0
.l
0
x
2F
CMJ
nV
L___
LO
71-
(I)
2.0
1.0
0.06
CM
Icl
X-ray dose (Gy)
Salpha particles
0C
cVJ
In
C,
0
0.1
0.6
T
T
1.2
6.0
!
Alpha particle dose (Gy)
Figure 3-14. Fraction (%) of the bystander AG01522 fibroblast cells showing
induction of y-H2AX foci after 4 hrs co-culture with irradiated DU-145 tumor cells.
The x-ray (3-14A) or alpha particle (3-14B) doses delivered to the DU-145 cells are
indicated on the x-axis. X-ray experiments (3-14A) were carried out with and without
the addition of DMSO or PTIO scavengers. Results are the average of at least four
independent experiments. Error bars represent + 1 SD. (** P < 0.01, * P < 0.05;
significance of scavengers were relative to the same absorbed dose delivered to the
directly-targeted cells in the absence of scavenger)
106
3.3 Cell Cycle Analysis of Bystander AG01522 Cells
y-H2AX foci induction is believed to be a marker for double strand breaks in
directly-irradiated cells. y-H2AX foci induction has also been shown to be expressed in
S-phase cells at stalled replication forks [118]. An increase of y-H2AX foci induction in
bystander AGO 1522 cells co-cultured with x-ray- irradiated DU-145 cells is shown in
Figure 3-14. Cell cycle analysis using Fluorescence Activated Cell Sorting (FACS) was
performed on the bystander AGO 1522 cells to test whether the detected y-H2AX foci
increase shown in Figure 3-14 was a result of a greater percentage of the cells entering Sphase. Dr. Martin Purschke generously performed the FACS analysis at Massachusetts
General Hospital. Figure 3-15 shows the cell cycle profile of bystander AGO 1522 cells 4
hrs after co-culture with unirradiated DU-145 cells. The percentage of bystander
AG01522 cells in each phase was: GI, 55.90%; S-phase, 26.29%; and G2, 17.58%. The
background level of apoptosis in the bystander AGO1522 cells was 0.19%.
I
IAUfRtme121.005
""~
Marke Evers %Gated %Total
AS 25676 100.00 65.59
M1
50
0.19
0.17
M2
14353
55.90 47.84
M3
s750
26.29 22.50
M4
4513
17.58 15.04
0
0
200
4
R2A
e"o
s
t1oo
Figure 3-15. FACS analysis of the bystander AG01522 fibroblast cells after 4 hrs of
co-culture with unirradiated DU-145 tumor cells.
107
Figure 3-16 shows the cell cycle profile of bystander AG01522 cells 4 hrs after co-culture
with DU-145 cells that had been irradiated with 2.0 Gy of x-rays. The percentage of
bystander AG01522 cells in each phase was: G1, 56.08%; S-phase, 27.72%; and G2,
16.03%. The apoptosis level was 0.21% in the bystander AG01522 cells co-cultured with
2.0 Gy x-ray-irradiated DU-145 cells.
I
Figure 3-16. FACS analysis of the bystander AG01522 fibroblast cells after 4 hrs of
co-culture with 2.0 Gy x-ray irradiated DU-145 tumor cells.
Two additional independent cell cycle analysis experiments were performed; both
confirmed the results shown in Figures 3-15 and 3-16. There was no significant increase
in the S-phase population of bystander AGO 1522 cells 4 hrs after co-culture with x-rayirradiated DU-145 cells as compared to the unirradiated co-cultured sample. Therefore,
the increase in y-H2AX foci induction in the bystander AG01522 cells could not be
attributed to an increase in the fraction of cells in S-phase. Yang et al. have also shown
that the y-H2AX induction seen in bystander AG01522 cells co-cultured with x-rayirradiated AG01522 cells was not due to an increase of S-phase bystander cells [64].
108
3.4 Survival Fraction of Bystander DU-145 Cells
The surviving fraction of bystander DU-145 cells was measured as a function of
increased dose delivered to the directly-irradiated tumor cells. Because previous MN
results showed that factor(s) in the bystander signaling cascade was (were) very short
lived in bystander effects with DU-145 cells (Section 3.1.1), the inserts containing the
DU-145 cells were added before the alpha particle irradiation. For x-ray irradiations, the
co-culture was initiated within 5 min after the exposure. No decrease in bystander cell
survival was seen with either of the radiation qualities when bystander DU-145 cells were
co-cultured with directly-irradiated DU-145 cells (Fig. 3-17).
2.0
0
1.0
> 0.9
() 0.8
0.7
Ipha particles
-rays
C
6.0
Dose (Gy)
Figure 3-17. Surviving fraction of bystander DU-145 cells following 4 hrs co-culture
with DU-145 cells that had been irradiated with either x-rays (A) or alpha particles
(*) at the doses indicated on the x-axis. Data points are the mean ± 1 SD of three
independent experiments for X-rays and four experiments for alpha particles.
109
3.5 Survival Fraction of Bystander AG01522 Cells
The surviving fraction of bystander AG01522 cells was measured as a function of
increased dose delivered to the directly-irradiated tumor cells. No decrease in survival
was seen in bystander AG01522 fibroblasts that were co-cultured with a-particle
irradiated DU-145 cells (Figure 3-18). The survival fraction of the bystander AGO1522
cells co-cultured with 0.1 Gy alpha-particle irradiated DU-145 cells is borderline
significant. This is not convincing since none of the other survival fraction data points
were significant for the bystander AG01522 cells co-cultured with alpha-particle
irradiated DU- 145 cells.
In contrast, a reduction to 80% survival fraction was seen in bystander AG01522
fibroblasts that were co-cultured with x-ray irradiated DU-145 cells (Figure 3-18). The
addition of PTIO blocked the survival fraction bystander effect, i.e., there was no
significant reduction in survival fraction in the fibroblasts that had been co-cultured with
x-ray-irradiated tumor cells (Figure 3-19). DMSO showed an intermediate effect in
terms of reducing the survival fraction bystander effect, but the differences relative to the
x-rays-alone group were not statistically significant (Fig. 3-19).
The data presented in Figures 3-16 represent a clear example of an LETdependent bystander effect: x-ray irradiation of DU-145 tumor cells produced a
significant reduction of survival fraction in AGO 1522 bystander fibroblasts, alpha particle
irradiation did not. In both cases, co-culture of the bystander AGO1522 cells with the
directly-irradiated DU-145 cells began within 5 min after irradiation.
110
CN
1.0
r-
O
0.9
0
cc
O
a- 0.8
LL
cles
"h._ 0.7
,
0.0
0.5
1.0
1.5
2.0
2.5
I
6.0
3.0
Dose (Gy)
Figure 3-18. Surviving fraction of bystander AG01522 cells following 4 hrs coculture with DU-145 cells that had been irradiated with either x-rays (A) or alpha
particles (e) at the doses indicated on the x-axis. Data points are the mean ± 1 SD
from four independent experiments for x-rays and two experiments for alpha particles.
(** P < 0.01 relative to the non-irradiated controls)
**
**
**
-~cL'
~
1
0
C
L_
LL
1.01
2i
0.9
I
0.8
S0.7
I
-
M X-rays + PTIO
0 X-rays + DMSO
A X-rays
0.6
0.0
I
I
I
I
0.5
1.0
1.5
2.0
Dose (Gy)
Figure 3-19. Surviving fraction of bystander AG01522 cells following 24 hrs coculture with DU-145 cells that had been irradiated with x-rays at the doses indicated
on the x-axis in the presence or absence of DMSO and PTIO scavengers. Data
points are the mean + 1 SD from four independent experiments with x-rays and five
experiments with scavengers. (** P < 0.01, * P < 0.05 relative to the no-scavenger-added
effect with x-rays at the same dose)
111
3.5.1 Temperature Effects on the Survival Fraction Bystander Signal
X-ray irradiations of the DU-145 cells at 40C were performed to further probe
whether the survival fraction bystander signal was chemical or biological in nature. The
surviving fraction of bystander AG01522 cells was measured as a function of increased
dose delivered to the tumor cells that were directly-irradiated at 40C. The survival
fraction of bystander AG01522 cells co-cultured with 250C x-ray-irradiated DU-145 cells
is shown in Figure 3-17 and is reproduced in Figure 3-18. In the 40C temperaturecontrolled experiments, the addition of the bystander AGO 1522 cells was delayed by an
additional 7 min in order to allow for the directly-irradiated cells to warm to 250C. This
was to ensure that the bystander cells were never added to cold medium. Irradiations of
DU-145 cells at 4oC significantly (P<.05) abolished the survival fraction bystander effect
in the co-cultured AG01522 fibroblasts (Figure 3-20). This indicates that the 40C x-ray
irradiation of the DU-145 cells prevented the release of the survival fraction bystander
signal, or some component of a signaling cascade.
The initiation of co-culture with DU-145 cells that were irradiated at 25 0C with xrays was also delayed in order to mimic the 40C temperature controlled experiments.
Figure 3-20 shows that the 7 min delay of the addition of AG01522 bystander cells to
25°C x-ray (2 Gy) irradiated DU-145 cells did not significantly change the survival
fraction response in the bystander cells. Therefore, the 250C directly-irradiated tumor
cells were either still emitting a bystander signal or the already-emitted signal was stable
in the medium during this delayed addition of the bystander cells. Therefore, the
bystander signal had not decayed prior to the delayed additions of the bystander cells for
the 4'C or 25 0C -controlled experiments.
112
*k
*l
1.1 -
4
MCC
*
25Cc
0
0 25 C- Delay
I
,)
13
S0.95
-
I
1
**
I
IF
I
I
0.9-
.
-W
~0.85U.
>• 0.80.750.7
0.7- -
I
0.1
I
I
0.5
1
2
Dose (Gy)
Figure 3-20. Surviving fraction of bystander AG01522 cells following 24 hrs coculture with DU-145 cells that had been irradiated with x-rays at (i) 40C, (ii) 25TC
with immediate co-culture initiation or (iii) 250C with a 7 min delay in co-culture
initiation. Data points are the mean ± 1 SD from at least four independent experiments.
(** P < 0.01, * P < 0.05 for 4oC effect with x-rays relative to the 25°C effect with x-rays
at the same dose)
3.6 Summary of Experimental Results
Table 3-2 summarizes all experiments in this thesis. The shaded regions
represent pairs of experiments that demonstrate different bystander responses due solely
to the different radiation qualities used to irradiate the DU-145 tumor cells (i.e., 250 kVp
x-rays vs. alpha particles). The two most clear-cut examples occur in bystander
AGO 1522 cells where significant y-H2AX and survival fraction bystander effects were
detected when the DU-145 cells were irradiated with x-rays, but no bystander effects
were detected after alpha particle irradiations. There were also LET-related differences
113
observed in the response of DU-145 bystander cells for the MN formation endpoint: the
signal produced by alpha particle-irradiated DU-145 cells was short-lived and PTIO did
not block this bystander effect; however, when the DU-145 target cells were irradiated
with x-rays, the signal was long-lived and PTIO completely eliminated the bystander
effect.
Bystander Effect
Endpoint
Bystander effect present? (Y/N)
DU-145 cells
DU-145 cells
directly irradiated
directly irradiated
with
with
Bystander cells
y-H2AX foci
NO
DU-145
NO
YES
YES
DMSO: blocks'
DMSO: blocks
AG01522
PTIO: blocks
Signal: long-lived
4oC - blocks
PTIO: blocks
Signal: long-lived
4oC - blocks
DU-145
NO
NO
MN formation
Survival Fraction
Table 3-2. Summary of bystander effect experiments with three cases (shaded cells)
of bystander effect differences due solely to radiation quality (LET). 1DMSO
appears to block the effect but the differences are not statistically significant.
114
Chapter 4
Discussion
The studies in this thesis have shown that irradiated DU-145 prostate carcinoma
cells can produce medium-mediated bystander effects in non-irradiated populations of
both DU-145 tumor cells and AG01522 normal human fibroblasts. The data indicate
that, in some situations, the bystander tumor cells and the bystander fibroblasts respond
differently to the same signal. Further, direct evidence of LET-dependent differences
was seen in the response of the bystander fibroblasts to the signals coming from the
irradiated tumor cells.
4.1 LET-dependence
The work presented in thesis has shown LET-dependent differences in the
response of the bystander AGO1522 cells to the signal coming from DU-145 cells
irradiated with x-rays or alpha particles. When bystander AG01522 cells were cocultured with x-ray-irradiated tumor cells, all three endpoints studied (y-H2AX, MN, and
survival fraction) exhibited bystander effects. In contrast, when the bystander AGO 1522
cells were co-cultured with alpha-particle irradiated tumor cells, only a MN bystander
effect was observed. The observance of a MN bystander effect in the AG01522 cells cocultured with alpha-particle-irradiated DU-145 cells confirmed that there was indeed a
115
bystander signal emitted. The bystander signal emitted from the irradiated tumor cells
was most likely composed of a cascade of factors. It can be hypothesized that certain
components of the bystander signal were responsible for an induction of one type of
detected damage, but were not responsible for an induction of a different type of detected
damage. The type of factors that are involved in the bystander-signaling pathway could
be LET-dependent. The factor(s) in the bystander-signaling cascade that had caused the
y-H2AX or survival fraction bystander effect in the AGO 1522 cells were a result of the xray irradiation but not the alpha-particle irradiation of the DU-145 cells.
When the bystander cells were the DU-145 cells instead of the AG01522 cells,
further LET-dependent differences were observed in the bystander response. Previous
work had shown that the bystander signal emitted from alpha-particle irradiated DU-145
cells was "short-lived" because the bystander DU-145 cells needed to be present in the
medium before/during irradiation to exhibit a MN bystander effect [10]. When the DU145 cells were irradiated with low-LET x-rays, the bystander cells were always added
within 5 min after the irradiation. The data presented in this thesis showed a significant
MN induction in the bystander DU-145 cells after co-culture with DU-145 cells irradiated
with x-rays. Therefore, the factor(s) in the bystander signal emitted from x-ray
irradiations of the DU-145 cells that induced a MN response in the bystander DU-145
cells was stable and "long-lived". This suggested that there were LET-dependent
differences in the life-time of the factor (or some component of the signal cascade) that
was critical for production of a MN response in bystander DU-145 cells.
Much of the bystander effect literature has been based on high-LET (mostly alpha
particle) irradiations for several reasons. Historically, early bystander effect reports used
116
low doses (particle fluences) of alpha particles from planar isotopic sources to generate a
mixed population of hit and non-hit cells based on hit probabilities defined by Poisson
statistics [40, 46]. The recent availability of accelerator-based microbeams capable of
delivering exact numbers of charged particles to individual cells, subcellular regions, or a
layer of tissue has made it possible to avoid Poisson statistics and the associated
uncertainties [8, 14, 76, 77, 120]. There are reports of bystander effects caused by lowLET radiation, both in medium-transfer [59, 89, 93, 94]and insert [43, 54, 63, 64]
experiments. There is surprisingly little information on the LET-dependence of
bystander effects carried out under conditions where all parameters remain the same and
only the LET is varied, as was done in the work presented in this thesis. What little
information that is available is contradictory.
Hickman et al. observed a greater-than-predicted fraction of rat lung epithelial
cells with increased p53 expression after exposure to a low fluence of alpha particles
[46]. When the cells were exposed to similar doses (< 10 cGy) of x-rays, there was no
increase in the fraction of cells showing increased p53 expression. While, technically,
this is an LET-dependence of the bystander response, it could also be a result of different
levels of signal released from the irradiated cells due to different damage thresholds for
the two radiation types, and/or the different fractions of cells actually traversed by
radiation and damaged for similar doses (related to the alpha particle Poisson hit
probabilities). Shao et al. reported an LET-dependent increase in MN induction in
bystander glioblastoma cells, with a greater effect produced by 100 keV/gim carbon ions
than 13 keV/pm carbon ions [15]. In the Shao study, there was no low-LET photon
control in those experiments to provide a true low-LET point of comparison. The same
117
group used a microbeam to target individual primary human fibroblasts with either 40Ar
(1260 keV/jtm) or 20Ne (380 keV/gm) ions [95]. The increase in MN observed in the cell
population was independent of the LET (particle) used or the number of particles
delivered to the targeted cells (1-4 particles per cell). This was reported as no LETdependence, but both of these particles have very high LET properties, and the ability of
the cell to respond may have been saturated (i.e., overkill). Lyng et al. have reported noLET dependent differences in bystander effects in a study intended to directly compare
medium-transfer experiments to a microbeam experiment [59, 89, 93]. When a proton
microbeam (3.2 MeV protons; - 11 keV/jtm) was used to target a population of human
keratinocytes growing on a mylar membrane in the same medium as a separate
unirradiated population, the degree of bystander response was similar to that with the
gamma-irradiated (0.2 keV/jtm) medium transfer method. These authors concluded that
"the mechanisms involved are common across different radiation qualities and
conditions", but the range of radiation quality tested was limited (0.2 - 11 keV/jtm) and
much lower than typical alpha particle LET values (-150 keV/jtm), where the majority of
the bystander effect literature has been generated. In contradiction to their own previous
conclusion, when the keratinocytes were irradiated with 1.3 mGy to 1.0 Gy doses of
neutrons there was no detection of a medium-mediated bystander effect in the
keratinocytes [94]. The experimental protocol used for both the neutron and the gammaray experiments were carried out in the exact same medium transfer conditions. The data
demonstrated a clear LET dependence of the bystander effect observed in the
keratinocytes. Boyd et al. used an external gamma-ray beam or pharmaceuticals tagged
with either a low-LET emitter (beta particle) or a high-LET emitter (auger and alpha-
118
particle emitter) to study medium-mediated bystander effects in two transfected human
tumor cell lines [96]. Bystander cells that received medium from high-LET irradiated
cells exhibited a U-shaped survival fraction curve. A steep reduction in survival fraction
was detected in the bystander cells at low radiation activities of high-LET particles.
Irradiations of cells with high activity concentrations of the high-LET particles caused the
survival fraction of the bystander cells to recover to near control levels. Bystander cells
that received medium from the low-LET irradiated cells exhibited reduction in survival
fraction at all measured doses or activities. The authors speculated that perhaps exposure
of the tumor cells to high activity concentrations of the high-LET radionuclides could
inhibit the cellular ability to generate bystander signals. Their study showed that the lowLET radiation qualities were more effective at causing cell death in the bystander tumor
cells than the high-LET particles.
The report concerning LET-dependence of the bystander effect that is most
directly comparable to the results reported in this thesis, is the work of Yang et al. using
250 kVp x rays (2 keV/gm) or 1 GeV/n Fe ions (151 keV/prm) to directly irradiate
AG01522 human fibroblasts and then monitor bystander effects in a separate population
co-cultured in the same medium on an insert [43, 63, 64]. The bystander effect endpoints
were y-H2AX formation, MN induction and a decrease in survival fraction of the
bystander cells. For all endpoints, these authors report no bystander effect differences
between the 1 GeV/n Fe ions and the 250 kVp x-rays, concluding that, in their system the
bystander effect was independent of LET. The results in this thesis differ, in that a
tumor cell population was irradiated. The endpoints used to assay effects in the bystander
AG01522 cells used in this thesis were the same as those in the Yang et al. studies [63,
119
64]. This suggests that the LET-dependence of the bystander effect in the AG01522 cells
is also dependent on the irradiated cell line.
4.2 The Nature of the Bystander Signal
For alpha particles, the average doses quoted here, and used in all of the Figures,
are the product of the particle fluence and the LET. Thus, the dose is an average value
delivered to the entire volume penetrated by the alpha particles. Yet on the cell level, any
individual cell is either hit or not hit. Those cells that are not hit, receive no dose. Those
cells that are directly hit receive a relatively high dose i.e., - 0.5 Gy. E.g., for the
irradiation geometry used for these studies, an alpha particle dose of 0.6 Gy represents a
mean of 3.4 alpha particles per nucleus with only 3% of the cells estimated to receive no
hits based on Poisson statistics [10]. For the same total dose, with x-rays there are many
electrons (from photo-electric and Compton scattering events) per cell, all contributing to
a relatively uniform energy deposition in all cells in the population. All cells would be
considered to be directly traversed by irradiation with x-rays.
Bystander effects involve two discrete steps: the production of a signal by directly
irradiated cells, and the response of the bystander cells to that signal. The nature of the
signal released into the medium by irradiated cells remains unknown. The experiments
reported here clearly show an LET-dependent difference in the response of AGO 1522
cells to the signal released by irradiated DU-145 cells. For both the y-H2AX and the
survival fraction endpoints, there was a significant bystander effect in the AGO 1522 cells
following x-ray irradiation of the DU-145 cells, but not when the DU-145 cells were
irradiated with alpha particles. This raises several possibilities regarding the nature of the
120
signal: 1) there is an LET-dependent difference in the amount of signal produced by the
DU-145 cells; 2) the high-LET alpha particle damage was severe, and the hit cells
produced no signal; or 3) there is a different signal released from irradiated DU-145
tumor cells depending on whether they are irradiated with x-rays or alpha particles.
Regarding possibilities 1 and 2), the different natures of the radiations used produce
profound differences in the dose delivery at the cellular level. At very low alpha particle
dose, there is significant fraction of cells not hit, and thus unlikely to produce a bystander
signal. Other reports have shown a threshold dose under which no bystander effects were
detected [46, 94]. The experiments in this thesis were designed with doses to the directly
irradiated cells that result in plateau levels of bystander effects in an effort to avoid this
type of problem: low percentages of cells hit at low doses creating "artifact" LET
differences, that in fact are nothing more than threshold differences due to the different
natures of the radiations used and different fractions of cells irradiated, or differences in
the amount of signal produced. The effects reported in this thesis occurred at relatively
high doses of alpha particles delivered to the directly irradiated cells, i.e. all cells should
be hit. As for the possible explanation that the irradiated DU-145 cells emitted no signal
after x-ray irradiation, this was not true because a MN bystander effect was observed in
the AGO 1522 cells co-cultured with x-ray-irradiated DU-145 cells.
Additional insight as to the nature of the bystander signal emitted from the
irradiated DU- 145 cells is provided by examination of the results shown in Table 3-2
regarding the ROS dependencies in the AG01522 bystander effects observed. The
bystander-signaling cascade had NO as a factor for the endpoints of y-H2AX, survival
fraction and MN in the bystander AG01522 cells co-cultured with DU-145 cells
121
irradiated with either alpha particles or x-rays. The bystander signaling cascade also had
ROS as a factor for the endpoint of y-H2AX in the bystander AGO1522 cells for both
radiation qualities. Although the results were not statistically significant, the data were
suggestive that ROS was present in the signaling for both survival fraction and MN
bystander effects in the AG01522 cells. Yang et al. have shown NO and ROS dependent
MN and y-H2AX inductions in bystander AGO1522 cells co-cultured with AGO1522 cells
irradiated with either x-rays or iron particles [43, 64]. In their studies, the survival
fraction bystander effect in the AG01522 cells was independent of ROS. While in these
published studies the directly irradiated cells were not the DU-145 cells, the bystander
AGO 1522 fibroblasts also responded to NO and ROS dependent bystander signals.
ROS dependencies in the bystander DU-145 cells were also observed. Previous
published results had shown that ROS-dependent, but NO-independent, bystander
signaling caused MN induction in DU-145 cells co-cultured with alpha-particle irradiated
DU-145 cells [10]. Shao et al. showed an NO-dependent MN induction in both human
glioblastoma (T98G) cells and in AG01522 bystander cells co-cultured with T98G cells
irradiated with alpha-particle microbeams [8, 14]. The Shao et al. results contradict the
data presented in this thesis. The NO-dependency of the bystander signal may be
influence by the tumor cell line used. Data presented in this thesis showed both ROS and
NO-dependent bystander signaling for MN induction in bystander DU-145 cells cocultured with x-ray irradiated DU-145 cells. ROS and NO production have been shown to
lead to MAPK activation. The MAPK signaling cascade has been shown to be involved
in bystander effects in both normal and tumor cells [9, 59, 88, 121]. Hagan et al. showed
significant MAPK activation through EGFR and TGF-a signaling in bystander DU-145
122
cells that received ICCM from gamma-ray irradiated DU-145 cells. Using a different
tumor cell line, Facoetii et al. showed significant releases of IL-8 into the medium from
directly irradiated tumor cells. IL-8 turned on the CXCRI receptors in the bystander cells
[91]. IL-8 has been shown to bind to a membrane receptor that initiates the MAPK.
Regulation of IL-8 expression in some human cells has been shown to be NO dependent
[122]. Regulation of COX-2 expression has also been shown to be dependent on nitric
oxide synthase [123]. NAD(P)H oxidase proteins, which have also been implicated in
bystander effects, are found in lipid rafts [15, 90]. NAD(P)H oxidase proteins have been
shown to be activated by NO and H202 exposure, which can be created by -OH [15, 124].
NAD(P)H oxidase proteins once activated also trigger ROS production. It can therefore
be postulated that the ROS production by NAD(P)H in turn activates the MAPK and
eventually the COX-2 pathways in bystander cells.
This study has shown ROS and NO dependence of the bystander effect in both
bystander AGO 1522 and DU-145 cells co-cultured with DU-145 cells irradiated with
either alpha-particles or x-rays. The scavengers were present in the medium that was
shared by both the directly-irradiated and bystander cell populations during the entire coculture time. Thus, it was not possible to determine whether the ROS or NO production
was due to the signal emitted from the directly irradiated cells or from the response in the
bystander cells. Enzymatic processes are severely slowed down when cells are held at
low-temperatures [99]. Physical damage in directly-irradiated cells has been shown to be
independent of the temperature at the time of the irradiation [37]. Therefore, it can be
assumed that the initial levels of DNA damage caused in the directly-irradiated DU-145
cells was the same regardless of the irradiation temperature. Hypothermic irradiation
123
conditions (40C) were used to study only the emission of the bystander signal from the of
DU-145 cells. This was accomplished by holding only the tumor cells in hypothermic
conditions prior to and during irradiation. Further, hypothermic conditions were able to
test whether the emitted bystander signal at 25TC was one of chemical or biological
nature. Alpha-particle irradiations of DU-145 cells at 40C were shown to block the MN
bystander effect in the bystander AGO 1522 cells. Similarly, 40C x-ray irradiations of the
DU-145 cells abolished the survival fraction and MN bystander effect in the AG01522
cells. X-ray irradiations of the DU-145 cells at 4TC, however, did not abolish the MN
bystander effect in the DU-145 cells. This suggested that the production of the factors
involved in the bystander signal emitted from irradiated DU-145 cells was both biological
(i.e. metabolic or protein-like) and chemical in nature. Furthermore, the data suggested
that the bystander response in the AGO 1522 cells was dependent on the biological factors
of the bystander signal emitted from either x-ray- or alpha-particle- irradiated DU-145
cells. In contrast, the bystander response in the DU-145 cells was dependent on the
chemical factors of the bystander signal emitted from the x-ray-irradiated DU-145 cells.
Mothersill et al. have reported that when irradiations of their keratinocytes were
performed at 4°C, no bystander signal was emitted [38]. The authors also concluded that
the bystander signal, or the initiation of the bystander signaling process must involve a
protein-like factor.
4.3 Different Cell Lines Respond Differently to the Same Signal
The bystander effect literature is full of seemingly contradictory reports of effects,
or lack of effects. However, most involve some differences in the experimental
124
conditions such as the use of different cell lines, where it is perhaps not surprising that
the response is different. The data presented here represent direct evidence for this: an
example of different cell lines responding differently to the same signal. With alphaparticle-irradiated DU-145 cells as the signal donor, both AG01522 cells and DU-145
cells showed a MN bystander effect. PTIO blocked the MN bystander effect in AG01522
cells, but had no effect on the MN bystander effect in DU-145 cells. In addition, the
factors that induced MN in the AGO 1522 cells appeared to be long-lived. In contrast, the
factors that induced MN in the DU-145 cells appeared to be short lived. Yet the signal
was the same (released from alpha-particle-irradiated DU-145 cells), representing a
different response of these two different cell lines to a common signal.
Some interesting patterns appear when the responses of the different cell lines are
examined for all three-bystander effect endpoints (see Table 3-2). When the DU-145
cells were directly irradiated with x-rays, the bystander AG01522 cells showed positive
bystander effects for all three endpoints, y-H2AX foci, MN formation, and a decrease in
survival fraction. In contrast, in response to the same signal (x-rays on DU-145 tumor
cells), the bystander DU-145 cells showed an increase in MN, but no increase in y-H2AX
foci, and no decrease in survival fraction. This raises the question of linkage between
these different endpoints in the two different cell lines. It has been shown that MN are
formed from unrepaired double strand breaks [50]. To further complicate the situation,
when the bystander signal originated from alpha particle-irradiated DU-145 cells, both
the bystander DU-145 tumor cells and the bystander fibroblasts showed the latter pattern:
MN formation, but no increase in y-H2AX foci, and no decrease in survival fraction.
The fact that the bystander DU-145 cells showed no increase in y-H2AX foci under any
125
conditions led us to try direct irradiation (see Section 3.2), to confirm that these tumor
cells were able to assemble y-H2AX foci. This was, indeed the case. DU-145 tumor cells
showed a 10-fold increase in y-H2AX foci induction after irradiation with either 10.0 Gy
of x-rays or 6.0 Gy of alpha particles. Macphail et al. showed a 10-fold induction and a
20-fold induction of y-H2AX foci in irradiated DU-145 cells with eitherl0.0 Gy or 50.0
Gy of x-rays, respectively [119]. This study confirmed the y-H2AX induction data for xray-irradiated DU-145 cells as reported by the Macphail group. In the AG01522
fibroblasts directly irradiated with 2 Gy of x-rays, 70% of the population showed
increased y-H2AX foci induction [43]. This suggests a difference in the doses required to
initiate y-H2AX foci formation between AG01522 fibroblasts and DU-145 tumor cells.
Using the same coverslip methodology as in this study, Burdak-Rothkamm et al. also
showed y-H2AX induction in normal human astrocytes that were co-cultured with human
glioma cells that had been irradiated with x-rays [16]. This group also observed a yH2AX foci induction in the human glioblastoma cells that were co-cultured with the
irradiated normal human astrocytes with x-rays. These data contradict the results seen in
this study, where y-H2AX induction was not detected in the bystander tumor cells. This
could possibly be attributed to differences in threshold doses required for initiation of yH2AX foci between the DU-145 and the human glioma cells.
The bystander effect is often described as a low dose effect, meaning that the
increase in biological damage seen in bystander cells is comparable in magnitude to the
damage seen in directly irradiated cells in the low dose range. For the DU-145 cells the
maximum MN bystander effect was equivalent to direct irradiation with an alpha particle
dose of 0.06 Gy [10]. For AGO 1522 cells, the magnitude of a y-H2AX foci formation
126
bystander effect was equivalent to direct irradiation with 0.1 Gy of 250 kVp x-rays [43].
In this thesis work, the DU-145 cells did not exhibit a significant increase in y-H2AX foci
formation until after 10 Gy of direct x-ray irradiation or 6 Gy of direct alpha particle
irradiation. In the low dose region, the directly irradiated DU-145 cells do not recruit the
y-H2AX histone protein for possible double strand break repair at significant levels. Why
both cell lines show the curious pattern of MN formation, but no increase in y-H2AX
foci, and no decrease in survival fraction when the signal originated from alpha particle
irradiated DU-145 tumor cells in unclear, but is likely related to LET-dependent
differences in the signal released, or to threshold differences in the ability of the different
cell lines to respond to these signals, or both.
127
Chapter 5
Summary & Conclusion
In summary, the data presented in this thesis indicates that:
(i) There is an LET-dependent difference in the signal emitted from the directly
irradiated DU-145 cells that causes medium-mediated bystander effects in both tumor
cells and normal human fibroblasts.
(ii) Tumor cells and fibroblasts can exhibit different bystander effect responses to
the same signal.
(iii) The bystander effect is NO and ROS dependent.
(iv) The initial emission of the bystander signal is dependent on both a proteinlike factor and a chemical factor from the irradiated DU-145 cells with x-rays. For alphaparticle irradiated DU-145 cells, the emission of the bystander signal is dependent on a
protein-like factor.
Whether the bystander effect is a biological protective mechanism or one of
enhancement of cell death is unknown. One possibility is that directly-irradiated cells
emit a bystander signal to neighboring cells in order to communicate that there has been
some sort of local damage and that additional nearby healthy cells must also be killed in
order to make room for a new healthy cell population. It is also possible that the
128
bystander signal emitted from the directly-irradiated cells specifically seeks to remove
neighboring unirradiated cells that might already be considered "damaged" or
"unhealthy" to the population. Another possibility is that the bystander signal itself is
purely toxic in nature and is randomly removing additional unirradiated cells. An
example of this beneficial/harmful mechanism is the observation that additional levels of
apoptosis are measured in bystander cells. Apoptosis can be viewed as a beneficial
removal of both abnormal directly-irradiated or bystander cells from a tissue population.
Apoptosis can be seen as a protective measure against pre-malignant responses. If there
are too many additional cells undergoing apoptosis induction due to cellular damage
caused by bystander effects, this can lead to breakdown of tissue function.
One commonality in bystander studies is that the effect is detected at very low
doses (<0.5 Gy), where radiation did not traverse the majority of the population. The
enhancement of genomic instability, mutation, and enhanced growth in bystander cells is
a concern for radiation protection. The bystander effect actually was first noticed in an in
vitro study that was set up to explore mutation frequencies from a radon-type of
exposure. At these very low doses, the concern with the bystander effect is that cancer
induction risks are actually higher than what is currently estimated using the linear no
threshold model for radiation risk. Understanding the mechanisms and importance of
bystander effects at low dose exposures to normal tissues can give a better perspective for
setting radiation exposure limits.
In the case of tumor therapy, the bystander effect could be responsible for two key
things: enhancement of tumor cell death and or additional cell death in nearby normal
tissues. If bystander effects are determined to be occurring in tumor cells in vivo, possible
129
modulation of the signal could be employed. If directly irradiated tumors emit a signal
that affects neighboring normal tissues in vivo, then it might be possible to manipulate
this signal to decrease this additional damage to normal cells. If an in vivo signal from
directly-irradiated tumor cells also increases damage in unirradiated neighboring tumor
cells, it would be advantageous if this signal could also be manipulated to enhance
overall tumor cell death. Depending on the cell specific bystander effect mechanism, the
bystander pathway could therefore either be inhibited or enhanced via drugs during tumor
therapy.
Both the in vitro and the in vivo reports on the bystander effect are contradictory.
The in vitro bystander studies are limited by nature since they can only focus at an
isolated component of what is a complicated signaling system response in vivo. This
limitation may explain some of the contradictions that is prevalent in the bystander effect
literature. In vivo signaling systems are complex processes. Therefore it is not too
surprising that the in vivo bystander effect literature is also contradictory. The true
relevance, however, of the bystander effect to the radiation biology field will only be
determined in vivo.
5.1 Future Work
The temperature experiments blocked the bystander effect, indicating that a proteinlike factor is involved in the bystander signaling cascade. Future studies should focus on
further probing the nature of the bystander signal that is emitted from the irradiated DU145 cells. The MAPK pathway has already been shown to be activated in bystander DU145 cells that were co-cultured with gamma-ray irradiated DU-145 cells [9]. The MAPK
130
pathway is initiated by binding of IL-8, IGF, TNF-c, and TNF-r to the outer cell
membrane. Levels of these proteins could be measured in the medium of the DU-145
cells following irradiation with either x-rays or alpha-particles. Antibodies specific to
these binding factors could be added prior to irradiation of the DU-145 cells to test
whether they abolish the bystander effect in either the tumor or normal cells. Activation
of the NADP(H) oxidase proteins has also been shown to be a factor in bystander studies.
NADP(H) oxidase proteins have also been shown to induce production of ROS. Future
studies can look for activation of NADP(H) proteins in the bystander DU-145 or
AG01522 cells as a possible explanation for the ROS dependent bystander effect
observed.
Bystander effect studies with DU-145 cells should also be expanded from the in vitro
studies to in vivo models. Initial experiments could be performed in analogous 3-D
tissues such as the EFT-300 and the Air-100, both of which are commercially available.
These were described in Section 1.2.2.2. The 3-D tissues could be developed to
incorporate DU-145 into their cellular matrix. Using the same approach as Sedelnikova et
al. [75], the bottom of the tissue could be irradiated with alpha particles. Bystander
effects, in the form of y-H2AX induction and MN formation, could then be detected in
both the normal and tumor cells at incremental distances away from the irradiated
surface. Gap junction inhibitors, such as lindane, could also be used to determine if GJIC
formed between the DU-145 cells plays a role in the release of a bystander signal.
Bystander effect experiments with DU-145 cells could be performed in mice. DU145 cells could be injected subcutaneously into a nude mouse. The tumor could then be
selectively irradiated in vivo using a targeted beam of charged particles, such as protons
131
or alpha particles. Or a 2 10Po needle (an alpha-particle source) could also be used to
irradiate only the tumor cells. Biological endpoints such as y-H2AX foci induction and
MN formation could be assayed as a function of distance away from the irradiated tumor
cells. Both unirradiated tumor cells and normal tissue cells at distances away from the
irradiated tumor cells could be assayed for bystander effects.
132
Bibliography
1.
Prise, K.M., M. Folkard, V. Kuosaite, L. Tartier, N. Zyuzikov, and C. Shao, What
rolefor DNA damage and repairin the bystander response? Mutat Res, 2006.
597(1-2): p. 1-4.
2.
3.
Schwartz, J.L., Variability: the common factor linking low dose-induced genomic
instability, adaptationand bystander effects. Mutat Res, 2007. 616(1-2): p. 196200.
Little, J.B., Cellularradiationeffects and the bystander response. Mutat Res,
2006. 597(1-2): p. 113-8.
4.
Little, J.B., LauristonS. Taylor lecture: nontargetedeffects of radiation:
implicationsfor low-dose exposures. Health Phys, 2006. 91(5): p. 416-26.
5.
Hamada, N., H. Matsumoto, T. Hara, and Y. Kobayashi, Intercellularand
intracellularsignalingpathways mediating ionizing radiation-inducedbystander
6.
effects. J Radiat Res (Tokyo), 2007. 48(2): p. 87-95.
Matsumoto, H., N. Hamada, A. Takahashi, Y. Kobayashi, and T. Ohnishi,
7.
Vanguards of paradigm shift in radiationbiology: radiation-inducedadaptive
and bystander responses. J Radiat Res (Tokyo), 2007. 48(2): p. 97-106.
Chaudhry, M.A., Bystander effect: biologicalendpoints and microarrayanalysis.
8.
Mutat Res, 2006. 597(1-2): p. 98-112.
Shao, C., M. Folkard, B.D. Michael, and K.M. Prise, Targeted cytoplasmic
irradiationinduces bystander responses. Proc Natl Acad Sci U S A, 2004.
101(37): p. 13495-500.
9.
Hagan, M., L. Wang, J.R. Hanley, J.S. Park, and P. Dent, Ionizing radiationinduced mitogen-activatedprotein (MAP) kinase activation in DU145 prostate
carcinoma cells: MAP kinase inhibition enhances radiation-inducedcell killing
and G2/M-phase arrest.Radiat Res, 2000. 153(4): p. 371-83.
10.
Wang, R. and J.A. Coderre, A bystandereffect in alpha-particleirradiationsof
human prostate tumor cells. Radiat Res, 2005. 164(6): p. 711-22.
11.
Xue, L.Y., N.J. Butler, G.M. Makrigiorgos, S.J. Adelstein, and A.I. Kassis,
Bystander effect produced by radiolabeledtumor cells in vivo. Proc Natl Acad Sci
12.
U S A, 2002. 99(21): p. 13765-70.
Kishikawa, H., K. Wang, S.J. Adelstein, and A.I. Kassis, Inhibitoryand
stimulatory bystander effects are differentially induced by Iodine-125 and Iodine-
123. Radiat Res, 2006. 165(6): p. 688-94.
13.
14.
Mothersill, C. and C. Seymour, Medium from irradiatedhuman epithelial cells
but not human fibroblasts reduces the clonogenic survival of unirradiatedcells.
Int J Radiat Biol, 1997. 71(4): p. 421-7.
Shao, C., M. Folkard, B.D. Michael, and K.M. Prise, Bystander signaling between
glioma cells andfibroblasts targeted with counted particles. Int J Cancer, 2005.
116(1): p. 45-51.
133
15.
Shao, C., Y. Furusawa, M. Aoki, H. Matsumoto, and K. Ando, Nitric oxidemediated bystander effect induced by heavy-ions in human salivary gland tumour
16.
cells. Int J Radiat Biol, 2002. 78(9): p. 837-44.
Burdak-Rothkamm, S., S.C. Short, M. Folkard, K. Rothkamm, and K.M. Prise,
ATR-dependent radiation-inducedgamma H2AX foci in bystanderprimary human
astrocytes and glioma cells. Oncogene, 2007. 26(7): p. 993-1002.
17.
Sutherland, B.M., P.V. Bennett, J.C. Sutherland, and J. Laval, ClusteredDNA
damages induced by x rays in human cells. Radiat Res, 2002. 157(6): p. 611-6.
18.
Bennett, P.V., N.S. Cintron, L. Gros, J. Laval, and B.M. Sutherland, Are
endogenous clustered DNA damages induced in human cells? Free Radic Biol
19.
Med, 2004. 37(4): p. 488-99.
Georgakilas, A.G., P.V. Bennett, D.M. Wilson, 3rd, and B.M. Sutherland,
Processingof bistrandedabasicDNA clusters in gamma-irradiatedhuman
20.
hematopoieticcells. Nucleic Acids Res, 2004. 32(18): p. 5609-20.
Paul, S., L. Gros, J. Laval, and B.M. Sutherland, Expression of the E. colifpg
protein in CHO cells lowers endogenous oxypurine clustered damage levels and
decreasesaccumulation of endogenousHprt mutations. Environ Mol Mutagen,
2006. 47(5): p. 311-9.
22.
Kulkami, A. and D.M. Wilson, 3rd, The involvement of DNA-damage and -repair
defects in neurologicaldysfunction. Am J Hum Genet, 2008. 82(3): p. 539-66.
Friedberg, E.C., A briefhistory of the DNA repairfield. Cell Res, 2008. 18(1): p.
23.
3-7.
Weinfeld, M., A. Rasouli-Nia, M.A. Chaudhry, and R.A. Britten, Response of
21.
base excision repairenzymes to complex DNA lesions. Radiat Res, 2001. 156(5 Pt
2): p. 584-9.
24.
Friedberg, E.C., How nucleotide excision repairprotects againstcancer. Nat Rev
25.
Cancer, 2001. 1(1): p. 22-33.
Kobayashi, J., K. Iwabuchi, K. Miyagawa, E. Sonoda, K. Suzuki, M. Takata, and
H. Tauchi, CurrentTopics in DNA Double-StrandBreak Repair.J Radiat Res
(Tokyo), 2008.
26.
Moore, J.K. and J.E. Haber, Cell cycle and genetic requirements of two pathways
of nonhomologous end-joining repairof double-strandbreaks in Saccharomyces
cerevisiae. Mol Cell Biol, 1996. 16(5): p. 2164-73.
27.
Lavin, M.F. and S. Kozlov, ATM activation and DNA damage response. Cell
28.
Cycle, 2007. 6(8): p. 931-42.
Rigberg, D.A., T.A. Blinman, F.S. Kim, M.A. Cole, and D.W. McFadden,
29.
30.
31.
Antisense blockade of p21/WAF1 decreasesradiation-inducedG2 arrestin
esophagealsquamous cell carcinoma.J Surg Res, 1999. 81(1): p. 6-10.
Storer, J.B. and L.H. Hempelmann, Hypothermia and increasedsurvival rate of
infant mice irradiatedwith x-rays. Am J Physiol, 1952. 171(2): p. 341-8.
Hornsey, S., Protectionfrom whole-body x-irradiationafforded to adult mice by
reducing the body temperature. Nature, 1956. 178(4524): p. 87.
Weiss, L., Some effects of hypothermia and hypoxia on the sensitivity of HeLa
cells to x-rays. Int J Radiat Biol, 1960. 2: p. 20-7.
32.
Smith, L.H. and T.W. McKinley, Jr., Influence of temperature on x-ray sensitivity
of mouse bone marrow cells in vitro. Radiat Res, 1967. 32(3): p. 441-51.
134
33.
34.
35.
36.
Belli, J.A. and F.J. Bonte, Influence of temperature on the radiationresponse of
mammalian cells in tissue culture. Radiat Res, 1963. 18: p. 272-6.
Cater, D.B. and L. Weiss, Measurements of oxygen tension in the tissues of mice
cooled to 1 degree C. Nature, 1959. 183(4674): p. 1521-2.
Elmroth, K., L.J. Erkell, and R. Hultborn, Influence of temperature on radiationinduced inhibition of DNA supercoiling. Radiat Res, 1999. 152(2): p. 137-43.
Claesson, A.K., B. Stenerlow, L. Jacobsson, and K. Elmroth, Relative biological
effectiveness of the alpha-particleemitter (211)Atfor double-strandbreak
induction in human fibroblasts.Radiat Res, 2007. 167(3): p. 312-8.
37.
Larsson, D.E., S. Gustavsson, R. Hultborn, J. Nygren, U. Delle, and K. Elmroth,
38.
Chromosomal damage in two X-ray irradiatedcell lines: influence of cell cycle
stage and irradiationtemperature.Anticancer Res, 2007. 27(2): p. 749-53.
Mothersill, C. and C.B. Seymour, Cell-cell contact during gamma irradiationis
not requiredto induce a bystandereffect in normal human keratinocytes:
evidence for release during irradiationof a signal controlling survival into the
39.
40.
41.
medium. Radiat Res, 1998. 149(3): p. 256-62.
Azzam, E.I., S.M. de Toledo, T. Gooding, and J.B. Little, Intercellular
communication is involved in the bystanderregulationof gene expression in
human cells exposed to very low fluences of alpha particles.Radiat Res, 1998.
150(5): p. 497-504.
Nagasawa, H. and J.B. Little, Induction of sister chromatid exchanges by
extremely low doses of alpha-particles.Cancer Res, 1992. 52(22): p. 6394-6.
Kashino, G., K.M. Prise, G. Schettino, M. Folkard, B. Vojnovic, B.D. Michael, K.
Suzuki, S. Kodama, and M. Watanabe, Evidencefor induction of DNA double
strandbreaks in the bystanderresponse to targetedsoft X-rays in CHO cells.
42.
Mutat Res, 2004. 556(1-2): p. 209-15.
Azzam, E.I., S.M. de Toledo, A.J. Waker, and J.B. Little, High and lowfluences
of alpha-particlesinduce a G1 checkpoint in human diploidfibroblasts.Cancer
Res, 2000. 60(10): p. 2623-31.
43.
44.
Yang, H., N. Asaad, and K.D. Held, Medium-mediated intercellular
communication is involved in bystander responses of X-ray-irradiatednormal
humanfibroblasts.Oncogene, 2005. 24(12): p. 2096-103.
Suzuki, M., H. Zhou, C.R. Geard, and T.K. Hei, Effect of medium on chromatin
damage in bystandermammalian cells. Radiat Res, 2004. 162(3): p. 264-9.
45.
Moore, S.R., S. Marsden, D. Macdonald, S. Mitchell, M. Folkard, B. Michael,
D.T. Goodhead, K.M. Prise, and M.A. Kadhim, Genomic instabilityin human
lymphocytes irradiatedwith individualchargedparticles: involvement of tumor
necrosisfactor alpha in irradiatedcells but not bystander cells. Radiat Res, 2005.
46.
163(2): p. 183-90.
Hickman, A.W., R.J. Jaramillo, J.F. Lechner, and N.F. Johnson, Alpha-particleinducedp53 protein expression in a rat lung epithelial cell strain. Cancer Res,
1994. 54(22): p. 5797-800.
47.
Nagasawa, H. and J.B. Little, Unexpected sensitivity to the induction of mutations
by very low doses of alpha-particleradiation:evidencefor a bystander effect.
Radiat Res, 1999. 152(5): p. 552-7.
135
48.
49.
Azzam, E.I., S.M. de Toledo, and J.B. Little, Directevidencefor the participation
of gapjunction-mediatedintercellularcommunication in the transmissionof
damage signalsfrom alpha -particleirradiatedto nonirradiatedcells. Proc Natl
Acad Sci U S A, 2001. 98(2): p. 473-8.
Concin, N., M. Stimpfl, C. Zeillinger, U. Wolff, L. Hefler, J. Sedlak, S. Leodolter,
and R. Zeillinger, Role of p53 in G2/M cell cycle arrestand apoptosis in response
to gamma-irradiationin ovarian carcinomacell lines. Int J Oncol, 2003. 22(1): p.
50.
51-7.
Fimognari, C., S. Sauer-Nehls, H. Braselmann, and M. Nusse, Analysis of
51.
radiation-inducedmicronuclei by FISH using a combination of painting and
centromeric DNA probes. Mutagenesis, 1997. 12(2): p. 91-5.
Azzam, E.I., S.M. de Toledo, and J.B. Little, Oxidative metabolism, gapjunctions
and the ionizing radiation-inducedbystandereffect. Oncogene, 2003. 22(45): p.
7050-7.
52.
Bishayee, A., H.Z. Hill, D. Stein, D.V. Rao, and R.W. Howell, Free radicalinitiatedand gapjunction-mediatedbystandereffect due to nonuniform
distributionof incorporatedradioactivityin a three-dimensionaltissue culture
model. Radiat Res, 2001. 155(2): p. 335-44.
53.
Hu, B., W. Han, L. Wu, H. Feng, X. Liu, L. Zhang, A. Xu, T.K. Hei, and Z. Yu,
In situ visualization of DSBs to assess the extranuclear/extracellulareffects
induced by low-dose alpha-particleirradiation.Radiat Res, 2005. 164(3): p. 286-
54.
91.
Fournier, C., D. Becker, M. Winter, P. Barberet, M. Heiss, B. Fischer, J. Topsch,
55.
and G. Taucher-Scholz, Cell cycle-relatedbystander responses are not increased
with LET after heavy-ion irradiation.Radiat Res, 2007. 167(2): p. 194-206.
Little, J.B., Genomic instabilityand bystander effects: a historicalperspective.
56.
Oncogene, 2003. 22(45): p. 6978-87.
Prise, K.M., G. Schettino, M. Folkard, and K.D. Held, New insights on cell death
from radiationexposure. Lancet Oncol, 2005. 6(7): p. 520-8.
57.
Hei, T.K., R. Persaud, H. Zhou, and M. Suzuki, Genotoxicity in the eyes of
bystander cells. Mutat Res, 2004. 568(1): p. 111-20.
58.
Ballarini, F., M. Biaggi, A. Ottolenghi, and 0. Sapora, Cellularcommunication
and bystander effects: a critical review for modelling low-dose radiationaction.
59.
60.
61.
Mutat Res, 2002. 501(1-2): p. 1-12.
Lyng, F.M., P. Maguire, B. McClean, C. Seymour, and C. Mothersill, The
involvement of calcium andMAP kinase signalingpathways in the production of
radiation-inducedbystander effects. Radiat Res, 2006. 165(4): p. 400-9.
Iyer, R. and B.E. Lehnert, Alpha-particle-inducedincreasesin the radioresistance
of normal human bystandercells. Radiat Res, 2002. 157(1): p. 3-7.
Mitchell, S.A., S.A. Marino, D.J. Brenner, and E.J. Hall, Bystander effect and
adaptive response in C3H O1T(1/2) cells. Int J Radiat Biol, 2004. 80(7): p. 465-
62.
72.
Hu, B., L. Wu, W. Han, L. Zhang, S. Chen, A. Xu, T.K. Hei, and Z. Yu, The time
and spatialeffects of bystanderresponse in mammalian cells induced by low dose
radiation.Carcinogenesis, 2006. 27(2): p. 245-51.
136
63.
Yang, H., V. Anzenberg, and K.D. Held, Effects of heavy ions and energetic
protons on normal humanfibroblasts. Radiats Biol Radioecol, 2007. 47(3): p.
302-6.
64.
65.
Yang, H., V. Anzenberg, and K.D. Held, The time dependence of bystander
responses induced by iron-ion radiationin normal human skin fibroblasts.Radiat
Res, 2007. 168(3): p. 292-8.
Ponnaiya, B., G. Jenkins-Baker, D.J. Brenner, E.J. Hall, G. Randers-Pehrson, and
C.R. Geard, Biological responses in known bystander cells relative to known
microbeam-irradiatedcells. Radiat Res, 2004. 162(4): p. 426-32.
66.
Scholz, M., W. Kraft-Weyrather, S. Ritter, and G. Kraft, Cell cycle delays
induced by heavy ion irradiationof synchronous mammalian cells. Int J Radiat
Biol, 1994. 66(1): p. 59-75.
67.
Tsuboi, K., Y. Tsuchida, T. Nose, and K. Ando, Cytotoxic effect of accelerated
carbon beams on glioblastomacell lines with p53 mutation: clonogenic survival
and cell-cycle analysis. Int J Radiat Biol, 1998. 74(1): p. 71-9.
68.
Schettino, G., M. Folkard, K.M. Prise, B. Vojnovic, K.D. Held, and B.D. Michael,
Low-dose studies of bystander cell killing with targetedsoft X rays. Radiat Res,
69.
2003. 160(5): p. 505-11.
Schettino, G., M. Folkard, B.D. Michael, and K.M. Prise, Low-dose binary
behaviorof bystandercell killing after microbeam irradiationof a single cell with
70.
focused c(k) x rays. Radiat Res, 2005. 163(3): p. 332-6.
Wu, L.J., G. Randers-Pehrson, A. Xu, C.A. Waldren, C.R. Geard, Z. Yu, and T.K.
Hei, Targetedcytoplasmic irradiationwith alphaparticles induces mutations in
71.
mammalian cells. Proc Natl Acad Sci U S A, 1999. 96(9): p. 4959-64.
Little, J.B., H. Nagasawa, G.C. Li, and D.J. Chen, Involvement of the
nonhomologous end joining DNA repairpathway in the bystander effect for
chromosomal aberrations.Radiat Res, 2003. 159(2): p. 262-7.
72.
Han, J., M.J. Hendzel, and J. Allalunis-Turner, Quantitativeanalysis reveals
asynchronous and more than DSB-associatedhistone H2AX phosphorylation
after exposure to ionizing radiation.Radiat Res, 2006. 165(3): p. 283-92.
73.
Pilch, D.R., O.A. Sedelnikova, C. Redon, A. Celeste, A. Nussenzweig, and W.M.
Bonner, Characteristicsof gamma-H2AXfoci at DNA double-strandbreaks sites.
74.
Biochem Cell Biol, 2003. 81(3): p. 123-9.
Medvedeva, N.G., I.V. Panyutin, I.G. Panyutin, and R.D. Neumann,
Phosphorylationof histone H2AX in radiation-inducedmicronuclei. Radiat Res,
75.
2007. 168(4): p. 493-8.
Sedelnikova, O.A., A. Nakamura, O. Kovalchuk, I. Koturbash, S.A. Mitchell,
S.A. Marino, D.J. Brenner, and W.M. Bonner, DNA double-strandbreaksform in
bystandercells after microbeam irradiationof three-dimensionalhuman tissue
76.
models. Cancer Res, 2007. 67(9): p. 4295-302.
Smilenov, L.B., E.J. Hall, W.M. Bonner, and O.A. Sedelnikova, A microbeam
study of DNA double-strandbreaks in bystanderprimary humanfibroblasts.
Radiat Prot Dosimetry, 2006. 122(1-4): p. 256-9.
77.
Sokolov, M.V., L.B. Smilenov, E.J. Hall, I.G. Panyutin, W.M. Bonner, and O.A.
Sedelnikova, Ionizing radiationinduces DNA double-strandbreaks in bystander
primary human fibroblasts.Oncogene, 2005. 24(49): p. 7257-65.
137
78.
79.
Rothkamm, K. and M. Lobrich, Evidencefor a lack of DNA double-strandbreak
repairin human cells exposed to very low x-ray doses. Proc Natl Acad Sci U S A,
2003. 100(9): p. 5057-62.
Rogakou, E.P., D.R. Pilch, A.H. Orr, V.S. Ivanova, and W.M. Bonner, DNA
double-strandedbreaks induce histone H2AX phosphorylationon serine 139. J
80.
Biol Chem, 1998. 273(10): p. 5858-68.
Desai, N., E. Davis, P. O'Neill, M. Durante, F.A. Cucinotta, and H. Wu,
Immunofluorescence detection of clustered gamma-H2AXfoci induced by HZEparticle radiation.Radiat Res, 2005. 164(4 Pt 2): p. 518-22.
81.
Balajee, A.S., B. Ponnaiya, R. Baskar, and C.R. Geard, Induction of replication
protein A in bystandercells. Radiat Res, 2004. 162(6): p. 677-86.
82.
Mothersill, C., R.J. Seymour, and C.B. Seymour, Bystander effects in repair-
83.
deficient cell lines. Radiat Res, 2004. 161(3): p. 256-63.
Nagasawa, H., L. Huo, and J.B. Little, Increasedbystander mutagenic effect in
DNA double-strandbreak repair-deficientmammalian cells. Int J Radiat Biol,
2003. 79(1): p. 35-41.
84.
Nagasawa, H. and J.B. Little, Bystander effect for chromosomal aberrations
induced in wild-type and repairdeficient CHO cells by low fluences of alpha
particles. Mutat Res, 2002. 508(1-2): p. 121-9.
85.
86.
Huo, L., H. Nagasawa, and J.B. Little, HPRT mutants induced in bystander cells
by very low fluences of alphaparticles resultprimarilyfrom point mutations.
Radiat Res, 2001. 156(5 Pt 1): p. 521-5.
Zhang, Y., J. Zhou, K.D. Held, R.W. Redmond, K.M. Prise, and H.L. Liber,
Deficiencies of double-strandbreak repairfactors and effects on mutagenesis in
directly gamma-irradiatedand medium-mediated bystander human
87.
lymphoblastoid cells. Radiat Res, 2008. 169(2): p. 197-206.
Ponnaiya, B., G. Jenkins-Baker, A. Bigelow, S. Marino, and C.R. Geard,
Detection of chromosomal instability in alpha-irradiatedand bystander human
fibroblasts.Mutat Res, 2004. 568(1): p. 41-8.
88.
Zhou, H., V.N. Ivanov, J. Gillespie, C.R. Geard, S.A. Amundson, D.J. Brenner, Z.
Yu, H.B. Lieberman, and T.K. Hei, Mechanism of radiation-inducedbystander
effect: role of the cyclooxygenase-2 signalingpathway. Proc Natl Acad Sci U S
A, 2005. 102(41): p. 14641-6.
89.
Maguire, P., C. Mothersill, C. Seymour, and F.M. Lyng, Medium from irradiated
cells induces dose-dependent mitochondrialchanges and BCL2 responses in
unirradiatedhuman keratinocytes. Radiat Res, 2005. 163(4): p. 384-90.
90.
Azzam, E.I., S.M. De Toledo, D.R. Spitz, and J.B. Little, Oxidative metabolism
modulates signal transductionand micronucleusformation in bystander cells
from alpha-particle-irradiated
normal human fibroblastcultures. Cancer Res,
91.
2002. 62(19): p. 5436-42.
Facoetti, A., F. Ballarini, R. Cherubini, S. Gerardi, R. Nano, A. Ottolenghi, K.M.
Prise, K.R. Trott, and C. Zilio, Gamma ray-inducedbystander effect in tumour
glioblastomacells: a specific study on cell survival, cytokine release and cytokine
receptors. Radiat Prot Dosimetry, 2006. 122(1-4): p. 271-4.
138
92.
Iyer, R., B.E. Lehnert, and R. Svensson, Factorsunderlying the cell growthrelated bystander responses to alphaparticles.Cancer Res, 2000. 60(5): p. 1290-
93.
8.
Lyng, F.M., P. Maguire, N. Kilmurray, C. Mothersill, C. Shao, M. Folkard, and
94.
factorsfrom microbeam irradiatedcells. Int J Radiat Biol, 2006. 82(6): p. 393-9.
Liu, Z., C.E. Mothersill, F.E. McNeill, F.M. Lyng, S.H. Byun, C.B. Seymour, and
K.M. Prise, Apoptosis is initiatedin human keratinocytes exposed to signalling
W.V. Prestwich, A dose thresholdfor a medium transferbystander effect for a
95.
human skin cell line. Radiat Res, 2006. 166(1 Pt 1): p. 19-23.
Shao, C., Y. Furusawa, Y. Kobayashi, T. Funayama, and S. Wada, Bystander
effect induced by counted high-LETparticlesin confluent humanfibroblasts: a
96.
mechanisticstudy. Faseb J, 2003. 17(11): p. 1422-7.
Boyd, M., S.C. Ross, J. Dorrens, N.E. Fullerton, K.W. Tan, M.R. Zalutsky, and
R.J. Mairs, Radiation-inducedbiologic bystander effect elicited in vitro by
targetedradiopharmaceuticalslabeled with alpha-, beta-, and auger electron-
emitting radionuclides.J Nucl Med, 2006. 47(6): p. 1007-15.
97.
98.
Seymour, C.B. and C. Mothersill, Relative contributionof bystander and targeted
cell killing to the low-dose region of the radiationdose-response curve. Radiat
Res, 2000. 153(5 Pt 1): p. 508-11.
Belyakov, O.V., A.M. Malcolmson, M. Folkard, K.M. Prise, and B.D. Michael,
Directevidence for a bystandereffect of ionizing radiationin primary human
fibroblasts.Br J Cancer, 2001. 84(5): p. 674-9.
99.
Wheeler, K.T., R. Hickman, G.B. Nelson, S.K. Moore, and C.A. Wallen,
Relationshipbetween DNA damage, DNA repair,metabolic state and cell
100.
lethality. Radiat Environ Biophys, 1992. 31(2): p. 101-15.
Mothersill, C., F. Lyng, C. Seymour, P. Maguire, S. Lorimore, and E. Wright,
Geneticfactors influencing bystandersignaling in murine bladderepithelium
after low-dose irradiationin vivo. Radiat Res, 2005. 163(4): p. 391-9.
101.
Persaud, R., H. Zhou, S.E. Baker, T.K. Hei, and E.J. Hall, Assessment of low
linearenergy transfer radiation-inducedbystander mutagenesis in a threedimensional culture model. Cancer Res, 2005. 65(21): p. 9876-82.
102.
Persaud, R., H. Zhou, T.K. Hei, and E.J. Hall, Demonstrationof a radiationinducedbystander effect for low dose low LET beta-particles.Radiat Environ
103.
Biophys, 2007. 46(4): p. 395-400.
Belyakov, O.V., S.A. Mitchell, D. Parikh, G. Randers-Pehrson, S.A. Marino, S.A.
Amundson, C.R. Geard, and D.J. Brenner, Biological effects in unirradiated
human tissue induced by radiationdamage up to I mm away. Proc Natl Acad Sci
104.
U S A, 2005. 102(40): p. 14203-8.
Koturbash, I., R.E. Rugo, C.A. Hendricks, J. Loree, B. Thibault, K. Kutanzi, I.
Pogribny, J.C. Yanch, B.P. Engelward, and 0. Kovalchuk, Irradiationinduces
DNA damage and modulates epigenetic effectors in distant bystander tissue in
vivo. Oncogene, 2006. 25(31): p. 4267-75.
105.
Goyal, R., R. Reinhardt, and A. Jeltsch, Accuracy of DNA methylation pattern
preservationby the Dnmtl methyltransferase.Nucleic Acids Res, 2006. 34(4): p.
1182-8.
139
106.
Lorimore, S.A., J.M. McIlrath, P.J. Coates, and E.G. Wright, Chromosomal
instabilityin unirradiatedhemopoietic cells resultingfrom a delayed in vivo
bystander effect of gamma radiation.Cancer Res, 2005. 65(13): p. 5668-73.
107.
Watson, G.E., S.A. Lorimore, D.A. Macdonald, and E.G. Wright, Chromosomal
instability in unirradiatedcells induced in vivo by a bystander effect of ionizing
108.
radiation.Cancer Res, 2000. 60(20): p. 5608-11.
Lorimore, S.A., M.A. Kadhim, D.A. Pocock, D. Papworth, D.L. Stevens, D.T.
Goodhead, and E.G. Wright, Chromosomalinstability in the descendants of
unirradiatedsurviving cells after alpha-particleirradiation.Proc Natl Acad Sci
U S A, 1998. 95(10): p. 5730-3.
110.
Schulz-Ertner, D. and H. Tsujii, ParticleRadiation Therapy Using Proton and
Heavier Ion Beams. J Clin Oncol, 2007. 25(8): p. 953-964.
Brechbiel, M.W., Targetedalpha-therapy:past,present,future? Dalton Trans,
111.
2007(43): p. 4918-28.
Zalutsky, M.R., D.A. Reardon, O.R. Pozzi, G. Vaidyanathan, and D.D. Bigner,
112.
Nucl Med Biol, 2007. 34(7): p. 779-85.
Metting, N.F., A.M. Koehler, H. Nagasawa, J.M. Nelson, and J.B. Little, Design
113.
of a benchtop alphaparticleirradiator.Health Phys, 1995. 68(5): p. 710-5.
Steiner, M.G. and C.F. Babbs, Quantitationof the hydroxyl radicalby reaction
114.
with dimethyl sulfoxide. Arch Biochem Biophys, 1990. 278(2): p. 478-81.
Repine, J.E., O.W. Pfenninger, D.W. Talmage, E.M. Berger, and D.E. Pettijohn,
109.
Targeted alpha-particleradiotherapywith 211At-labeled monoclonal antibodies.
Dimethyl sulfoxide prevents DNA nicking mediated by ionizing radiationor
iron/hydrogenperoxide-generatedhydroxyl radical.Proc Natl Acad Sci U S A,
115.
1981. 78(2): p. 1001-3.
Maeda, H., T. Akaike, M. Yoshida, and M. Suga, Multiplefunctions of nitric
oxide in pathophysiology and microbiology: analysis by a new nitric oxide
scavenger.J Leukoc Biol, 1994. 56(5): p. 588-92.
116.
117.
Fenech, M. and A.A. Morley, Cytokinesis-block micronucleus method in human
lymphocytes: effect of in vivo ageing and low dose X-irradiation.Mutat Res,
1986. 161(2): p. 193-8.
Hamada, N., G. Schettino, G. Kashino, M. Vaid, K. Suzuki, S. Kodama, B.
Vojnovic, M. Folkard, M. Watanabe, B.D. Michael, and K.M. Prise, Histone
H2AX phosphorylationin normal human cells irradiatedwithfocused ultrasoftX
rays: evidence for chromatin movement during repair.Radiat Res, 2006. 166(1 Pt
1): p. 31-8.
118.
119.
Ward, I.M. and J. Chen, Histone H2AX is phosphorylatedin an ATR-dependent
manner in response to replicationalstress. J Biol Chem, 2001. 276(51): p. 47759-
62.
MacPhail, S.H., J.P. Banath, T.Y. Yu, E.H. Chu, H. Lambur, and P.L. Olive,
Expression ofphosphorylatedhistone H2AX in culturedcell lines following
exposure to X-rays. Int J Radiat Biol, 2003. 79(5): p. 351-8.
120.
Hall, E.J., The bystander effect. Health Phys, 2003. 85(1): p. 31-5.
140
121.
Hwang, Y.S., M. Jeong, J.S. Park, M.H. Kim, D.B. Lee, B.A. Shin, N. Mukaida,
L.M. Ellis, H.R. Kim, B.W. Ahn, and Y.D. Jung, Interleukin-1betastimulates IL8 expression through MAP kinase and ROS signalingin human gastric carcinoma
122.
cells. Oncogene, 2004. 23(39): p. 6603-11.
Xiong, Q., Q. Shi, X. Le, B. Wang, and K. Xie, Regulation of interleukin-8
expression by nitric oxide in human pancreaticadenocarcinoma.J Interferon
123.
Cytokine Res, 2001. 21(7): p. 529-37.
Landino, L.M., B.C. Crews, M.D. Timmons, J.D. Morrow, and L.J. Marnett,
Peroxynitrite,the coupling product of nitric oxide and superoxide, activates
prostaglandinbiosynthesis. Proc Natl Acad Sci U S A, 1996. 93(26): p. 15069-74.
124.
Li, W.G., F.J. Miller, Jr., H.J. Zhang, D.R. Spitz, L.W. Oberley, and N.L.
Weintraub, H(2)0(2)-induced 0(2) production by a non-phagocyticNAD(P)H
oxidase causes oxidant injury. J Biol Chem, 2001. 276(31): p. 29251-6.
141