Lasers in Surgery and Medicine 43:152–163 (2011)
In Vitro Investigation of the Biological Effects
Associated With Human Dermal Fibroblasts Exposed to
Gerald J. Wilmink,1,2* Benjamin D. Rivest,1Caleb C. Roth,3Bennett L. Ibey,1Jason A. Payne,1
Luisiana X. Cundin,4Jessica E. Grundt,1Xomalin Peralta,5Dustin G. Mixon,1and William P. Roach1
1711th Human Performance Wing, Radio Frequency Radiation Branch, Air Force Research Laboratory,
Brooks City-Base, Texas
2National Academy of Sciences, NRC Research Associate Program, Washington D.C.
3General Dynamics Advanced Information Services, San Antonio, Texas
4Conceptual Mindworks, Inc, San Antonio, Texas
5University of Texas at San Antonio, San Antonio, Texas
Background: Terahertz (THz) radiation sources are
increasingly being used in military, defense, and medical
applications. However, the biological effects associated
with this type of radiation are not well characterized.
In this study, we evaluated the cellular and molecular
response of human dermal fibroblasts exposed to THz
Methods: In vitro exposures were performed in a temper-
ature-controlled chamber using a molecular gas THz laser
(2.52THz, 84.8mWcm?2, durations: 5, 10, 20, 40, or
80minutes). Both computational and empirical dosimetric
techniques were conducted using finite-difference time-
domain (FDTD) modeling approaches, infrared cameras,
and thermocouples. Cellular viability was assessed using
conventional MTT assays. In addition, the transcriptional
activation of protein and DNA sensing genes were eval-
uated using qPCR. Comparable analyses were also con-
ducted for hyperthermic and genotoxic positive controls.
38C during all THz exposures. We also found that for each
exposure duration tested, the THz and hyperthermic
exposuregroups exhibited equivalent levels of cellsurvival
(?90%) and heat shock protein expression (?3.5-fold
increases). In addition, the expression of DNA sensing
and repair genes was unchanged in both groups; however,
appreciable increases were observed in the genotoxic
Conclusions: Human dermal fibroblasts exhibit compa-
rable cellular and molecular effects when exposed to THz
radiation and hyperthermic stress. These findings suggest
that radiation at 2.52THz generates primarily thermal
effects in mammalian cells. Therefore, we conclude that
THz-induced bioeffects may be accurately predicted with
conventional thermal damage models. Lasers Surg. Med.
? 2010 Wiley-Liss, Inc.
The Terahertz (THz) region of the electromagnetic (EM)
spectrum is defined as frequencies ranging from 0.1 to
10THz. Historically, fewer sources have been available for
the THz region than for neighboring infrared (IR) and
microwave spectral bands; however, over the past decade,
numerous sources have been developed [1–22]. These
solutions to a host of basic research problems. In addition,
THz sources are increasingly being integrated into many
medical, military, and security applications. For instance,
cancer diagnosis [23–32], to recognize targets at distance
[33–35], and to identify concealed explosives, drugs, and
prompted concerns regarding the hazards and biological
effects associated with this type of radiation. These
concerns are primarily because few THz bioeffects studies
have been conducted. In fact, most of these studies were
conducted for the same international project—the ‘‘THz-
BRIDGE’’ project . The goals of this project were to
examine the mechanisms governing THz interactions with
cells and biomolecules, and to assess the genotoxicity of
THz radiation. Most studies associated with this project
via photo-thermal mechanisms (i.e., thermal effects). This
Gerald J. Wilmink and Benjamin D. Rivest shared first author.
Contract grant sponsor: AFRL, HQAF SGRS Clinical Investi-
gation Program Project title: Determination of Cellular Bioeffect
Thresholds for Terahertz Frequencies; Contract grant sponsor:
National Science Foundation (NSF-PREM); Contract grant
*Correspondence to: Gerald J. Wilmink, 8262 Hawks Road,
Brooks City Base, TX 78235-5128, USA.
Accepted 23 June 2010
Published online 25 August 2010 in wileyonlinelibrary.com.
? 2010 Wiley-Liss, Inc.
finding is not surprising because THz wavelengths are
known to be strongly absorbed by water (a?400cm?1at
2.52THz)—primary constituent of biological materials
. In contrast, mixed results were reported in regards
to the genotoxic effects of THz radiation [41–44]. For
instance, Korenstein-Ilan et al.  observed both geno-
toxic and epigenetic effects in lymphocytes exposed to
0.1THz radiation, whereas neither of these effects were
observed in several other comparable studies [41,44]. In
addition to affecting DNA, THz radiation has also been
shown to impair protein and enzymatic processes .
Overall, these findings are compelling because the energy
of THz photons is not high enough to directly disrupt
chemical bonds. Hence, the mechanism by which THz
radiation couples to, and subsequently damages biomole-
cules (e.g., DNA and proteins) remains unclear.
To date, empirical bioeffects studies have not been
conducted at 2.52THz. However, several theoretical mod-
els have been developed which suggest that this particular
THz frequency may couple directly to biomolecules via
coherent excitations  or linear/nonlinear resonance
mechanisms [47,48]. This coupling mechanism is believed
oscillates on the same time scale (picoseconds) as the
natural phonon frequencies of double-stranded DNA
(dsDNA) [47,48]. In addition to coupling, these models also
suggest that 2.52THz radiation can create localized open-
ings (bubbles) between the DNA strands [47,48]. These
openings subsequently drive the dsDNA to ‘‘unzip,’’ which
interferes with transcription processes. Overall, since
these mechanisms are extremely difficult to verify exper-
imentally, the concept of non-thermal or microthermal
effects remains a subject of constant debate.
In light of the above findings, a few fundamental
questions remain unanswered: first, do mammalian
cells exposed to 2.52THz radiation exhibit specific
cellular and/or molecular effects (e.g., protein and DNA
damage)? Second, are these effects comparable to those
observed in hyperthermic and genotoxic positive con-
trols? To answer these questions, we conducted in vitro
THz exposures on dermal fibroblasts using a high-power
THz laser tuned to a frequency of 2.52THz. To examine
the cellular and molecular effects, we then conducted
viability assays. In addition, to determine if protein and/
or DNA damage occurred, we used qPCR to quantify the
transcriptional activation of genes involved in protein
and DNA sensing and repair pathways. Finally, to
determine if these effects were THz-specific, we also
performed comparable analyses for hyperthermic and
genotoxic positive controls.
Cell Culture Conditions
Normal adult human dermal fibroblasts (HDF) were
cultured in Dulbecco’s modified Eagle’s medium and
supplemented with 10% fetal bovine serum. HDFs were
plated in 96-well polystyrene (PS) tissue culture-treated
plates in 100ml of media at a seeding density of 105cells/
well. Cells were incubated overnight and exposed the
next day, as previously described [49,50].
Custom-Designed Terahertz Exposure Enclosure
In order to conduct temperature-controlled in vitro THz
exposures, we designed an exposure system consisting of
gas purging system, operator control area equipped with
PVC gloves, input ports for both short- and long-THz
cavities, optical platform, and a sapphire window for IR
thermographic measurements (Fig. 1A). To ensure the
chamber was exhibiting consistent heat distribution,
spatial temperature measurements were conducted at
various positions in the enclosure. Measurements were
made using four Omega thermocouples affixed to a cross-
shaped holder made of PVC. Temperature readings were
collected from the bottom to the top of the chamber using
2cm increments (z-axis). The data showed that the
chamber maintained consistent temperatures (?0.58C) at
all positions .
For all measurements in this study, we used a stabilized,
integrated, far-IR, optically pumped molecular gas THz
laser source (SIFIR-50 OPTL, Coherent-DEOS, Santa
Clara, CA). This source is tunable across much of the THz
lines of ?100mW. The system uses a 50W tunable CO2
pump laser, a pump frequency reference lock, and two non-
folded THz laser cells—optimized for either short-wave or
long-wave operation. The short-wave cavity provides out-
put frequencies from 0.8 to 7THz, whereas the long-wave
cavity provides output frequencies from 0.3 to 7THz. The
CO2pump laser was used to excite the rotational bands of
methanol (CH3OH) gas. All experiments in this study were
conducted at a frequency of 2.52THz (l¼118mm). Optimal
performance was achieved using the following settings:
CH3OH gas pressure of 380mTorr.
Terahertz Beam Manipulation and Delivery Optics
Three optical elements were used to control and deliver
the THz beam from the source to the exposure plate (Fig.
1B,C). The first element is a microprocessor-controlled,
motor-driven shutter (Sutter Instr. Co, Novato, CA). This
shutter has a 25-mm diameter aperture and a 38milli-
to THz beam divergence, the shutter was positioned 7cm
from the opening of the short-wave cavity. Once delivered
through the shutteraperture,a flatgold mirror (d¼25mm)
and a parabolic silver mirror were used to collect and focus
radius (rb) was measured to be 3mm (Fig. 1C).
Terahertz Detectors and Beam Diagnostics
THz output power was measured with an Astral
Vector Series H410 calorimeter (Scientech, Boulder, CO).
This power meter covers laser power and energy measure-
TERAHERTZ BIOEFFECTS IN DERMAL FIBROBLASTS153
ments from 3mW to 30W and mJ to 30J, respectively. This
meter provides an absolute power reading (i.e., NIST-
traceable certification), which is required for standards
studies. Prior to conducting in vitro THz exposures, beam
diagnostic tests were conducted to ensure that the THz
source was generating consistent Gaussian shaped beam
profiles. Spatial beam profiles were evaluated using a
Spiricon Pyrocam III detector array. A sample image of the
THz beam profile is provided in Figure 1D (Spiricon Inc,
Computational dosimetry. Finite-difference time-
domain (FDTD) computational modeling techniques were
radiation. A two-dimensional homogeneous model was
used to predict the temperature rise over time within the
cellmedia.Themodel utilized afinite-difference numerical
technique to solve the heat equation using cylindrical
coordinates. With these approximations, the bioheat
equation can be written as:
where T(z,r,t) represents the temperature (K) as a function
of time and position, and A(z,r,t) is the power per unit
volume (Wm?3) absorbed within the media. This power is
provided by the THz source. The remaining terms define
the thermal properties of the cell media and are assumed
constant, with K being the thermal conductivity, c the
specific heat, and r the mass density of the media. For
these simulations, K was set to 0.5WK?1m?1, c to
3,900JK?1kg?1, and r to 1,000kgm?3. The heat equation
was solved numerically using an explicit finite-difference
scheme. The source term A(z,r,t) was evaluated using the
Beer–Lambert law . We assumed uniform absorption
in the radial direction and exponential decay in the axial
direction. The EM properties of the media were calculated
Fig. 1. Schematic representation of elements used for con-
trolled transmission, delivery, exposure, and evaluation of
cells irradiated with THz radiation. A: Macroscopic image of
in vitro THz exposure setup: molecular gas THz laser source,
CO2 Laser spectrometer, temperature-controlled exposure
chamber, and electric heater. B: Magnification of THz trans-
mission and delivery optics: electric shutter, flat gold plated
mirror, parabolic silver plated mirror, well plate holder
(adjustable in XYZ), and IR camera. C: Magnification of
exposure setup. D: Sample representative image of THz beam
profile at air–well interface measured with Pyrocam III
detector array. After accounting for losses in the polystyrene
plate, the power was calculated to be 24mW. Given a beam
radius of 3mm, the incident irradiance was found to equal
154WILMINK ET AL.
using a double-Debye fit, and were specified in terms of
relative permittivity (4.4) and conductivity (169.8Sm?1)
were used: (1) a convective boundary condition at air–
media interface (z¼Z) using a heat transfer coefficient (h)
at the bottom (z¼0) and outer walls of plate (r¼R). These
walls were assumed to be finite conductive layers with a
thickness of 1.5mm and a K of 0.08WK?1m?1. The outer
temperature was set to equal the ambient temperature of
378C. The geometry of the simulations was selected to
replicate the empirical exposure conditions (i.e., 100ml
media in a 6mm diameter cuvette). This volume corre-
as the z dimension in the model. The spatial resolution was
set to 0.01mm in both the r and z dimensions. Finite-
difference equations were iterated in time until an
asymptotic temperature profile was achieved. To compare
our predicted and empirically measured temperatures,
results were saved at predetermined intervals.
Empirical dosimetry. We used an IR camera and
thermocouples to measure the temperature of the cells
before, during, and after THz exposures. The IR camera
was placed 15cm from the media surface, and temperature
FLIR Systems Inc, Wilsonville, OR). Since IR camera
measurements only provide a measure of the surface
temperature, we also used thermocouples to measure
temperatures near the cell monolayer. The thermocouples
had the following specifications: T-type, Copper material,
0.00500diameter, Teflon insulation, and a response time of
40milliseconds (Omega Engineering, Inc., Stamford, CT).
trols. HDFs were exposed to 2.52THz radiation using an
irradiance of 84.8mWcm?2for 5, 10, 20, 40, or 80minutes.
For comparison, we also evaluated the response of hyper-
thermic and ultraviolet (UV) positive controls. The hyper-
40.08C for 5, 10, 20, 40, or 80minutes. This temperature
THz-exposed cells. The genotoxic, UV-exposed positive
controls, were exposed to a UV lamp (l¼254nm and 38W)
in a tissue culture hood for 3minutes. For these exposures,
uncovered plates were positioned 10cm from the lamp.
Cellular viability assays. Viability was evaluated
24hours post-exposure using MTT (3-(4,5-dimethylthia-
zol-2-yl)-2,5-diphenyltetrazolium bromide) assays, as per
manufacturer’s instructions (ATCC, Manassas, VA). In
stress,and were incubated for24hours. Then, the next day
we added 10ml of MTT reagent to the exposed wells, and
incubated them for 24hours until precipitate was visible.
We then added 100ml of detergent to each well, and left the
plate at room temperature for 2hours. Absorbance read-
ings were measured at 570nm with a Synergy HT Plate
in culture medium. Absorbance values were plotted versus
the cell number, and these curves were used to determine
the number of viable cells in each well.
Gene expression analyses. To determine whether
THz radiation induced direct damage to intracellular
the magnitude of expression for several signature gene
markers. For these experiments, RNA was harvested from
samples 4hours post-exposure using RNeasy Mini-Kits
(Qiagen, Valencia, CA). Previous studies showed that
HDFs exhibited maximal levels of mRNA expression at
a NanoDrop Spectrophotometer (NanoDrop Technologies,
BioanalyzerTM(Agilent Technologies, Santa Clara, CA).
greater than 9.5. PCR runs were performed on a StepOne-
PlusTMRT PCR system using TaqMan1Assays and
The following assays were used to evaluate protein
damage: HSP70 (HSPA1A, Hs00359163_s1; HSPA6,
Hs00275682_s1; HSPA4L, Hs00204666_m1); HSP40
(DNAJA4, Hs00388055_m1; DNAJB1, Hs00428680_m1;
m1); housekeeping gene (b-actin, ActB-Hs03023880_g1;
Life Technologies Corporation, Carlsbad, CA). Calibrator
RNA from HDFs was used as a control (Cell Applications
Inc., San Diego, CA).
To evaluate transcriptional activation of DNA sensing
for DNA Damage (Applied Biosystems, Foster City, CA).
This DNA Damage Induced 14-3-3 Sigma Signaling plate
measured the expression for the following DNA damage
probes and endogenous controls: 18S, GAPDH, HPRT1,
GUSB, ACTB, B2M, RPLP0, HMBS, TBP, PGK1, UBC,
PPIA, TBP, PGK1, TFRC, ATM, ATR, BRCA1, CCNB1,
CCNB2, CCNB3, CCNE1, CCNE1, CCNE2, CDC2, CDK2,
HUS1, RAD1, RAD17, RAD9A, SFN, TP53, YWHAE,
YWHAG, and YWHAQ. All PCR measurements were
conducted using a three-program LightCycler1protocol,
as previously described .
Predicting Temperature Profiles With
Computational Dosimetric Techniques
the temperatures generated in the THz-exposed cells.
Figure 2A contains a series of time-lapse images for the
temperature profiles predicted for a well of cells and media
exposed to THz radiation (H¼70.7mWcm?2). The data
?40.58C within the first minute of the exposure, whereas
the edge and the surface of the media only reach 39.5 and
388C, respectively. In addition, the data shows that the
images. This suggests that thermal equilibrium is reached
?3minutes into the exposure.
To examine therelationship
irradiance and temperature, we then used our model to
TERAHERTZ BIOEFFECTS IN DERMAL FIBROBLASTS 155
generate thermal history plots for several irradiances.
The majority of the irradiances were selected because
they corresponded to the range of irradiances that our
THz source could achieve (e.g., 42.4–100mWcm?2),
whereas a few others were included because they are
current international exposure limits (e.g., 1, 5, 10, and
100mWcm?2) [53–57]. Figure 2B is a plot of the temper-
ature–time curves for the media surface (z¼3.5mm)
generated with the model. For all irradiances tested,
the temperature of media rises sharply for the first
3minutes, and thereafter reaches a plateau. Maximum
and minimum temperatures of 40.2 and 38.48C were
observed using irradiances of 100 and 42.4mWcm?2,
7mWcm?2increase in irradiance, a 0.228C increase in
temperature is observed. Interestingly, the irradiances
which are suggested as exposure limits (e.g., 1, 5, and
10mWcm?2) increase the media temperature by less than
Since our experimental approach involves exposing a
monolayer of cells positioned at the well surface (z¼0) and
an IR camera provides surface temperature measure-
ments, we next used our computational model to predict
temperatures at various positions in the well. Figure 2C
presents the maximum temperatures at various depths in
an exposed well with media. As expected, the data shows
that the maximum and minimum temperatures are
observed at the well–media interface (z¼0) and the air–
media interface (z¼3.5mm), respectively. Interestingly, it
is also clear from the data that the relationship between
irradiance and temperature is not linear. For instance,
100mWcm?2achieves a temperature of 42.758C at z¼0
and 40.258C at z¼3.5mm (DTtop–bottom¼2.508C), whereas
42.4mWcm?2achieves a temperature of 39.508C at z¼0
and 38.258C at z¼3.5mm (DTtop–bottom¼1.258C). Finally,
for the irradiances currently suggested as exposure limits
(1, 5, and 10mWcm?2), the maximum temperatures are
only marginally above 378C.
Fig. 2. Computational dosimetry: using finite-difference time-domain (FDTD) models to
predict temperature profiles for cells exposed to THz radiation. A: Sequential time-lapse
images of the temperature profiles for cells in a well of media exposed to THz radiation
(u¼2.52THz, H¼70.7mWcm?2). B: Effect of irradiance on maximum temperature at
the media surface (z¼3.5mm). C: Maximum temperature at different locations within a
156 WILMINK ET AL.
Comparison of Temperature Images
Generated Using Computational and
Empirical Dosimetric Techniques
To test the fidelity of our computational models, we used
an IR camera to empirically measure the temperatures
generated in a well of media exposed to THz radiation
(u¼2.52THz, H¼84.8mWcm?2, t¼0–80minutes). Sam-
ple representative time-lapse images are provided in
Figure 3A. For the most part, the spatio-temporal temper-
ature profiles for the FDTD data (top panel) and IR camera
data (bottom panel) are in good agreement; however, a few
noticeable differences exist. First, the model predicts that
the temperature of the well rises more rapidly than we
observed experimentally. In fact, the model predicts that
the well reaches a final temperature of ?408C within
3minutes, whereas the empirical data requires much
more time to reach that temperature (Fig. 3A). The
second noticeable difference is that the model predicts that
the entire surface of the media surface will reach 408C (red
region), whereas the empirical data shows that only a
smaller region (radius ?1.5mm) actually reaches this
Quantitative Comparison of Computational
and Empirical Dosimetry
To determine the thermal history of cells exposed to THz
radiation, we used an IR camera and thermocouples to
monitor the temperatures before, during, after THz
exposures. In addition, we conducted experiments to
measure the temperature profiles for wells with and
without media to determine if the PS culture plate was
heating up during exposures. All thermal history data is
provided in Figure 3B. In brief, the data shows the
following: (1) both empirical techniques (thermocouple
and IR camera) measure comparable temperature rises of
?38C; (2) these temperatures were in good agreement with
those predicted with the computational model; (3) the
temperature of an empty well of a PS culture plate
increases by ?0.48C during exposures (Fig. 3B).
Specifically, we found that for a well with media, the
thermocouple, IR camera, and model data all show
comparable maximum temperatures: 40.25, 39.50, and
39.708C, respectively (Fig. 3B). However, although the IR
camera and thermocouple measure comparable temper-
atures, it is clear from the data that the thermocouple
measures slightly higher temperatures than the IR
camera. This is most likely due to the fact that it was
and thereby was warmer than the media surface, which is
the temperature the IR camera records. Overall, the
temperatures measured with the IR camera and thermo-
couple were comparable, and were in good agreement with
those predicted by the computational model.
Viability of Dermal Fibroblasts Exposed to
THz Radiation or Hyperthermic Stress
After conducting computational and empirical dosimet-
ric studies, we next sought to compare the viability of cells
exposed to THz radiation or hyperthermic stress. For these
studies, HDFs were exposed to THz radiation or hyper-
thermic stress (T¼40.08C) for 5, 10, 20, 40, or 80minutes.
For both treatment groups and untreated controls, we
Fig. 3. Comparison of computational and empirical dosimetric techniques. A: Time-lapse
images of temperature profiles for the surface of media in a well exposed to THz radiation.
The images in the top panel were collected with an IR camera for well of media exposed to
panel were generated using FDTD modeling techniques. B: Quantitative temperature-time
data collected for wells exposed to THz radiation using IR cameras and thermocouples.
Data were collected for wells with (w/) and without (w/o) media.
TERAHERTZ BIOEFFECTS IN DERMAL FIBROBLASTS 157
measured viability 24hours post-exposure using MTT
assays (Fig. 4A). The data shows the following: (1) the
(3) comparable levels of survivability are observed for both
THz Radiation Does Not Induce Direct Damage to
Previous studies have shown that THz radiation can
cause direct damage to intracellular proteins [45,47]. In
addition, it is known that cells respond to intracellular
protein damage by upregulating their gene expression for
several heat shock proteins (HSPs): Hsp40, Hsp70, and
Hsp105 [49,58–63]. Therefore, to determine if THz-
exposed cells exhibit elevated levels of protein damage—
compared to hyperthermic controls—we used HSP expres-
sion as a surrogate marker for protein damage. Gene
expression was quantified for the following HSPs: HSP70
(HSPA1A, HSPA4L), Hsp105 (HSPH1), and HSP40
fold-increase for each gene. For both sample groups, we
found that HSPA1A expression increases with exposure
duration, and is maximal (?2.5- to 3.2-fold increases) for
the 80minutes exposures (Fig. 4B). Interestingly, the
hyperthermic group shows slightly higher HSPA1A levels
than the THz-exposed group. For the other HSP genes
tested, we found that both groups show marginal increases
in expression (?1.8-fold; Fig. 4C–F). Of all the genes and
exposure conditions tested, we found that HSPH1 was the
Fig. 4. Comparison of cellular viability and gene expression
profiles for dermal fibroblasts exposed to THz radiation and
hyperthermic stress. Dermal fibroblasts were exposed to THz
radiation (u¼2.52THz, H¼84.8mWcm?2) or hyperthermic
stress (T¼408C) for 5, 10, 20, 40, or 80minutes. MTT viability
assays were conducted and qPCR was performed to measure
geneexpression changes. A: Averagenumber ofviable cells for
each group. Hyperthermic, gray bars; THz, black bars. Data
are expressed as means ? SD. B–F: Gene expression for
minimalstress proteins (HSPA1A,
DNAJA4, and DNAJB1) using qPCR. The mRNA expression-
fold values ð2???CtÞ were measured for sham and treatment
groups with n¼6 for each group. Values were calculated in
relation to b-actin and normalized to a separate RNA
calibrator. Data are expressed as mean ? SD; ***P<10?3,
**P<10?2, *P<0.05; between indicated groups.
158WILMINK ET AL.
only gene that exhibited higher increases in the THz-
exposed cells (e.g., 80minute exposure; Fig. 4D).
THz Radiation Does Not Activate DNA Sensing and
Recent theoretical models propose that 2.52THz radia-
tion may couple strongly to the natural breather modes of
DNA, thereby causing direct damage to such structures
[47,48]. To investigate if THz radiation affects DNA, we
conducted qPCR studies to measure the expression of DNA
reported that cyclin E (CCNE2) is a sensitive surrogate
marker for stress-induced DNA repair mechanisms, we
measured CCNE2 gene expression for cells exposed to THz
radiation, genotoxic stress (i.e., UV radiation), and hyper-
thermic-stress. The mean fold increases in CCNE expres-
that cells exposed to UV radiation—a stressor known to
induce DNA damage—do in fact exhibit a marked increase
(?40-fold) in CCNE2 expression. The data also shows that
the cells exposed to THz radiation and hyperthermic stress
do not upregulate this specific DNA sensing gene. In
addition to CCNE2, we also used a comprehensive DNA
numerous other well-characterized DNA sensing and
repair genes: HPRT1, GUSB, RPLP0, HMBS, TBP,
PGK1, UBC, PPIA, TFRC, ATM, ATR, BRCA1, CCNB1,
CCNB2, CCNE1, CCNE2, CDC2, CDK2, HUS1, RAD1,
RAD17, YWHAE, YWHAG, YWHAQ. We found that none
differences in gene expression compared to sham (data not
Given the strong absorption of water at THz frequencies
(a?400cm?1at 2.52THz), one would assume that the
biological effects associated with THz radiation are pri-
marily photothermal in nature . However, several
recent empirical studies have shown that THz radiation
can cause direct damage tointracellular biomolecules (e.g.,
lipids, proteins, and DNA) [42,43,45,66]. Since the irradi-
ances used in these studies only increased the temper-
that the preferential absorption of THz radiation by
biomolecules—via coherent excitations or linear/nonlinear
resonance mechanisms—may contribute to these observed
‘‘microthermal’’ effects [42,43,45–48,66]. Furthermore,
recent theoretical models postulate that these effects may
be pronounced at a frequency of 2.52THz [47,67]. Given
these findings, we sought to answer the following funda-
mental question: Do dermal fibroblasts exposed to high-
power 2.52THz radiation exhibit specific cellular and/or
molecular effects that are not observed in hyperthermic
and genotoxic controls?
The results of this study indicate that dermal fibroblasts
exposed to high-power THz radiation and hyperthermic-
stress exhibit comparable cellular and molecular effects.
Using both computational and empirical dosimetric tech-
niques, we provide evidence that the temperature of the
cells increases by ?38C during THz exposures. In addition,
of the THz-exposed cells survive (?90%), and comparable
levels are observed in the hyperthermic group. The qPCR
data show that HSPA1A levels are increased by ?3-fold for
suggests that THz radiation is not causing any additional
strain on intracellular proteins (Fig. 4B). Last, we show
that both the THz-exposed fibroblasts and the hyper-
thermic controls do not differentially express any of the
DNA repair and sensing genes; however, marked upregu-
lation of CCNE2 is observed in the genotoxic controls. In
summary, we show that high-power 2.52THz radiation
does not appear to cause direct damage to intracellular
biomolecules (lipid, proteins, DNA).
Since few studies have characterized the biological
effects associated with THz radiation, we created a table
to compare our results to those of previous studies (see
exposure parameters used in each study: frequency (THz),
delivery mode(CWvs.pulsed),exposureduration(ttot, min),
and incubation temperature during exposure (Texp). In
addition, we provide a detailed list of the bioeffects that
were observed: temperatures generated during exposure,
viability, changes in cell growth, and structural/functional
changes to lipid membranes, proteins, and DNA. It is clear
Fig. 5. THz radiation does not induce the activation of DNA
sensing and repair genes in dermal fibroblasts. Dermal
fibroblasts were exposed to UV radiation, a heated water bath
for 80minutes, or THz radiation for 80minutes. qPCR was
conducted on RNA extracted from samples 4hours post-
ð2???CtÞ are provided for Cyclin E2 (CCNE2). Values were
calculated in relation to b-actin and normalized to a separate
***P<10?3, **P<10?2, *P<0.05; between indicated groups.
expressedasmean ? SD;
TERAHERTZ BIOEFFECTS IN DERMAL FIBROBLASTS 159
from the data that the exposure parameters, cell type, and
endpoint evaluation techniques used in each study vary
significantly. Therefore, although it is difficult to draw
overarching conclusions, a few similar cellular and molec-
ular effects were observed in each study.
In most previous studies, THz radiation did not cause
changes at a cellular level (e.g., morphology, viability,
growth kinetics) [41,44,45,68]; however, a few studies
did observe such effects. Specifically, Olshevskaya et al.
 observed that neurons exhibited morphological
changes when exposed to 2.5THz radiation, and Kore-
nstein-Ilan et al. observed changes in growth kinetics for
cells exposed to THz radiation for long-exposure durations
(e.g., 1,440minutes) . In this study, we did not observe
any morphological changes in our THz-exposed cells, but
we did observe that the exposed fibroblasts exhibited
enhanced proliferation. A likelyexplanation for our results
is that the exposure conditions we used generated modest
the proliferation of cells [49,50]. In addition, for the longer
exposures, we also show that THz radiation can affect the
viability and growth of dermal fibroblasts; however, these
effects are most likely due to the fact that we used the
highest irradiance to date (e.g., ?2,500 times greater than
used in previous studies). Moreover, since our hyper-
thermic exposed cells also showed signs of death, these
findings further confirm that the THz-induced cell death
we observed is most likely mediated via photothermal
level, the data in Table 1 also suggests that mixed
conclusions have been drawn regarding the effects that
and DNA). For instance, Ramundo-Orlando et al. 
showed that pulsed THz radiation can directly affect the
permeability of liposomes. This observation is quite
intriguing, and it will be interesting to see if future
investigations observe similar effects in mammalian cells.
In addition to lipids, Homenko et al.  has also shown
that THz radiation can affect protein activity in an
antigen–antibody model; however, studies have yet to be
conducted to confirm if such effects are observed in
mammalian cells. In contrast, in this study we did not
observe appreciable increases in HSP expression, which
would presumably be elevated in cells exhibiting signifi-
cant intracellular protein damage.
Mixed conclusions have been drawn on the effect that
THz radiation has on DNA. For instance, both Scarfi et al.
and Zeniet al. showed that THzradiation doesnot
induce genotoxicity, while Korenstein-Ilan et al. 
observed signs of aneuploidy (i.e., genomic instability).
Although all of these studies used roughly the same
frequency (?0.1THz) and the same cell line (lymphocytes),
account for these conflicting results. First, Korenstein-Ilan
et al. exposed cells for considerably longer exposure
durations—1,440minutes rather than 20minutes.Second,
generates incoherent CW THz radiation, whereas Zeni
(FEL) that generates coherent, pulsed, broadband THz
radiation. To date, it is not clear which of these properties
may be responsible for these conflicting results.
Overall, although the exposure parameters and results
vary widely in the published THz studies, a few general
conclusions can be drawn. First, THz radiation can be
create considerable temperature increases [44,68].Second,
THz radiation can affect the growth and/or proliferation of
cells under certain exposure conditions [42,44,68]. Third,
pulsed THz radiation may be able to increase the perme-
ability of cellular membranes .
In addition to better understanding the bioeffects
associated with THz radiation, the findings of this study
may also be important for defining exposure standards at
TABLE 1. Summary of Cellular and Molecular Effects Observed in Previous THz Bioeffects Studies
(8C) % Alive
0.1CW 60 0.008R/TAB****?Recognition*
Homenko et al.
et al. 
Zeni et al. 
Scarfi et al. 
et al. 
et al. 
0.1 CW 14400.031 37L 0.3*þ**þAP
1.0–3.080MHz302.3 R/TK NSNSNSNS**
2.52CW8084.8 37F 3.090% NSNSþ HSPNS
*, Measurements not conducted in the study; NS, values with insignificant differences. AB, antibodies; L, lymphocytes; LP,
liposomes; K, keratinocytes; N, neurons; F, fibroblasts.
160 WILMINK ET AL.
THz frequencies. Historically, minimum safety standards
have been established to ensure that workers and the
general population are protected against adverse health
organizations function to define these standards, and
examples include: the European Parliament & Union (EP
& EU), International Commission on Non-Ionizing Radia-
tion Committee (ICNIRP), American National Standards
Institute (ANSI), and the Institute of Electrical and
Electronics Engineers (IEEE), European Committee for
Electrotechnical Standardization (CENELEC), Australia/
New Zealand (AUS/NZ), and The National Radiological
Protection Board (NRPB) [53,56,57].
is complicated for several reasons. First, each safety
organization uses different guidelines (e.g., scientific
rationale and policy judgments) to define their own safety
standard (see Fig. 6). Second, the THz frequency band
to both laser safety standards (0.3–10THz) and radio-
frequency (RF) radiation safety standards (0.1–0.3THz).
Third, laser and RF standards are defined using different
endpoint determinants, where the laser standard (ANSI
RF standard (IEEE C95.1) uses psychophysical perception
of pain and/or heating; Last, few empirical studies have
been conducted at THz frequencies, and as a result current
standards have been extrapolated from neighboring spec-
tral ranges. Given these above challenges, the empirical
data provided in this study may hopefully serve as a
starting point for defining a scientifically based exposure
limit fortheTHzband.It mayalsohelp facilitatethe global
harmonization of current standards .
In an effort to consider our results in the context of
current standards, we have provided a plot that summa-
rizes the exposure limits for most of the major health
100mWcm?2while ICNRIP defines it to be 10mWcm?2.
Since we show that the majority of cells exposed to
84.8mWcm?2survive, these data may suggest that an
exposure limit of 10mWcm?2may be conservative. In
addition, our computational modeling data also shows that
an irradiance of 10mWcm?2is predicted to increase the
temperature of water by only 0.48C over an 80-minute
period, whereas an irradiance of 100mWcm?2is predicted
to increase the temperature ?48C over the same time
period. In summary, we hope that the computational
models and empirical data collected in this study may help
in the development of empirically based standards for the
In this study, show that dermal fibroblasts exposed to
THz radiation (u¼2.52 THz, H¼84.8mWcm?2, and
t¼80minutes) do not exhibit appreciable changes at a
cellular and molecular level. Specifically, we show that
exhibit only minor increases in expression of HSPs. More-
increases are also observed in our hyperthermic controls.
Since the cells exposed to THz radiation did not express
higher levels of minimal stress proteins, this data suggests
intracellular proteins. Finally, our qPCR data also shows
that cells exposed to THz radiation do not express statisti-
cally significant changes in any of the DNA repair and
Research Associateship program and the Air Force
Research Laboratory for providing us with the opportunity
to conduct this study. In particular, we would like to thank
Frank Ruhr for his dedication and attention to detail in
design of the THz enclosure, and Dr. Morley Stone for his
consistent support. This work was supported by grants
provided by AFRL, HQAF SGRS Clinical Investigation
program: ‘‘Determination of Cellular Bioeffect Thresholds
for Terahertz Frequencies.’’ This work was also supported
by Partnership for Education and Research in Materials
(PREM) provided by the National Science Foundation
(NSF), Division of Materials Research Award (#0934218).
Fig. 6. International exposure standards at Terahertz fre-
quencies. The defined exposure limits (mWcm?2) are plotted
versus frequency (THz). The ‘‘worker’’ or controlled exposure
limit is denoted with a (O), and the general population
exposure limit is denoted with a (GP). Several data sources
were used in this plot [53,56,57]. EP & EU, European
Parliament & Union; ICNIRP, the International Commission
on Non-Ionizing Radiation Committee; AUS/NZ, Australia/
New Zealand; ANSI, the American National Standards
Institute; IEEE, the Institute of Electrical and Electronics
Engineers. The stardenotes the THzexposure conditions used
in this study.
TERAHERTZ BIOEFFECTS IN DERMAL FIBROBLASTS161
1. Ferguson B, Zhang X-C. Materials for terahertz science and
2. HosakoI, SekineN,PatrashinM,SaitoS, Fukunaga K,Kasai
Y, Baron P, Seta T, Mendrok J, Ochiai S, Yasuda H. At the
dawn of a new era in terahertz technology. Proc IEEE
3. Lee M, Wanke MC. Applied physics. Searching for a solid-
state terahertz technology. Science 2007;316(5821):64–65.
4. Siegel PH. Terahertz technology. Microwave theory and
techniques. IEEE Trans 2002;50(3):910–928.
5. Sirtori C. Quantum cascade lasers: Breaking energy bands.
Nat Photonics 2009;3(1):13–15.
6. WilliamsBS. Terahertz quantum-cascade
7. Ozyuzer L, Koshelev AE, Kurter C, Gopalsami N, Li Q,
TachikiM, KadowakiK, Yamamoto T,MinamiH, Yamaguchi
H, Tachiki T, Gray KE, Kwok WK, Welp U. Emission of
coherent THz radiation from superconductors. Science
8. Tonouchi M.Cutting-edge
9. Carr GL, Martin MC, McKinney WR, Jordan K, Neil GR,
Williams GP. High-power terahertz radiation from relativ-
istic electrons. 2002;420(6912):153–156.
10. Chen H-T, Padilla WJ, Cich MJ, Azad AK, Averitt RD, Taylor
AJ. A metamaterial solid-state terahertz phase modulator.
11. Mukherjee M, Mazumder N, Roy SK, Goswami K. GaN
IMPATT diode: A photo-sensitive high power terahertz
source. Semiconductor Sci Technol 2007;22(12):1258–1267.
12. Mukherjee M, Roy SK. Wide-Bandgap III–V nitride based
avalanche transit-time diode in Terahertz regime: Studies on
the effects of punch through on high frequency character-
istics and series resistance of the device. Curr Appl Phys
13. Buniatyan VV, Aroutiounian VM, Zekentes K, Camara N,
Soukiassian P. Silicon carbide TUNNETT diodes. Solid-State
14. Ryzhii V. Graphene tunneling transit-time terahertz oscil-
lator based on electrically induced p–i–n junction. Appl Phys
15. Nishizawa J, Kurabayashi T, Makabe H. 706-GHz GaAs CW
fundamental-mode TUNNETT diodes fabricated with molec-
ular layer epitaxy. Phys Status Solidi(c) 2008;5(9):2802–2804.
16. Asada M. Resonant tunneling diodes for sub-Terahertz and
Terahertz oscillators. Jpn J Appl Phys 2008;47:4375–4384.
17. Williams GP. High-power terahertz synchrotron sources.
Philos Transact A Math Phys Eng Sci 2004;362(1815):403–
413, discussion 413–414.
18. Grosse E. THz radiation from free electron lasers and its
potential for cell and tissue studies. Phys Med Biol
19. Wilmink GJ, Ibey BL, Roth CL, Vincelette RL, Rivest BD,
Horn CB,BernhardJ, RobersonD,RoachWP. Determination
of death thresholds and identification of terahertz (THz)-
specific gene expression signatures. In: Duco Jansen E,
Thomas RJ, editors. Optical interactions with tissues and
cells XXI. Bellingharn, WA, USA: Society of Photo-Optical
Instrumentation Engineers, Vol 7562, 2010. pp 75620K1–8.
20. Wilmink GJ, Rivest BD, Ibey BL, Roth CL, Bernhard J,
Roach WP. Quantitative investigation of the bioeffects
associated with terahertz radiation. In: Duco Jansen E,
Thomas RJ, editors. Optical interactions with tissues and
cells XXI. Bellingharn, WA, USA: Society of Photo-Optical
Instrumentation Engineers, Vol 7562, 2010. pp 75620L1–10.
21. Wilmink GJ, Roach WP. Measurement of the optical proper-
ties of skin using terahertz time-domain spectroscopic
techniques. In: Duco Jansen E, Thomas RJ, editors. Optical
interactions with tissues and cells XXI. Bellingharn, WA,
USA: Society of Photo-Optical Instrumentation Engineers,
Vol 7562, 2010. pp 75620J1–4.
22. Wade F, Smirnov D, Kumar SWS, HuQ. RL. Magnetic-field-
assisted terahertz quantum cascade laser operating up to
23. Oh SJ, Maeng I, Shin HJ, Lee J, Kang J, Haam S, Huh Y-M,
Suh J-S, Son. J-H. Nanoparticle contrast agents for terahertz
medical imaging. In Infrared, Millimeter and Terahertz
Waves. IRMMW-THz 2008. 33rd International Conference,
24. Oh SJ, Kang J, Maeng I, Suh JS, Huh YM, Haam S, Son JH.
Nanoparticle-enabledterahertz imagingfor cancer diagnosis.
Opt Expr 2009;17(5):3469–3475.
25. Wolbarst AB, Hendee WR. Evolving and experimental
technologies in medical imaging. Radiology 2006;238(1):
26. Bourne N, Clothier RH, D’Arienzo M, Harrison P. The effects
of terahertz radiation on human keratinocyte primary
cultures and neural cell cultures. Altern Lab Anim 2008;
27. Singh RS, Taylor ZD, Culjat MO, Grundfest WS, Brown
ER. Towards THz medical imaging; reflective imaging of
animal tissues. Stud Health Technol Inform 2008;132:472–
28. Suen JY, Tewari P, Taylor ZD, Grundfest WS, Lee H, Brown
ER,Culjat MO, SinghRS. Towardsmedical terahertzsensing
of skin hydration. Stud Health Technol Inform 2009;142:
29. WallaceVP,TadayPF,FitzgeraldAJ,WoodwardRM,Cluff J,
Pye RJ, Arnone DD. Terahertz pulsed imaging and spectro-
scopy for biomedical and pharmaceutical applications. Fara-
day Discuss 2004;126:255–263, discussion 303–311.
30. Wallace VP, Fitzgerald AJ, Shankar S, Flanagan N, Pye R,
Cluff J, Arnone DD. Terahertz pulsed imaging of basal cell
carcinoma ex vivo and in vivo. Br J Dermatol 2004;151(2):
31. Woodward RM, Cole BE, Wallace VP, Pye RJ, Arnone DD,
Linfield EH, Pepper M. Terahertz pulse imaging in reflection
geometry of human skin cancer and skin tissue. Phys Med
32. Woodward RM, Wallace VP, Pye RJ, Cole BE, Arnone DD,
Linfield EH, Pepper M. Terahertz pulse imaging of ex vivo
basal cell carcinoma. J Invest Dermatol 2003;120(1):72–78.
33. McClatchey K, Reiten MT, Cheville RA. Time resolved
synthetic aperture terahertz impulse imaging. Appl Phys
34. Heremans RV, Acheroy M. Synthetic aperture imaging
extended towards novel THz sensors. Sensors IEEE 2008;
35. Chan WL, Deibel J, Mittleman DM. Imaging with terahertz
radiation. Rep Prog Phys 2007;70(8):1325–1379.
36. Federici JF, Sehulkin B, Huang F, Gary D, Barat R, Oliveira
F, Zimdars D. THz imaging and sensing for security
applications: Explosives, weapons and drugs. Semiconductor
Sci Technol 2005;20(7):S266–S280.
37. Bogue R. Terahertz imaging: A report on progress. Sensor
38. Dobroiu A, Otani C, Kawase K. Terahertz-wave sources and
imaging applications. Meas Sci Technol 2006;17(11):R161–
39. Gallerano GP. THz-bridge: Terahertz radiation in biological
research, investigation on diagnostics and study of potential
genotoxic effects. Gian Piero Gallerano; 2004.
40. Segelstein D. The complex refractive index of water. Kansas
City, MO: University of Missouri; 1981.
41. Zeni O, Gallerano GP, Perrotta A, Romano M, Sannino A,
Sarti M, D’Arienzo M, Doria A, Giovenale E, Lai A, Messina
G, Scarfi MR. Cytogenetic observations in human peripheral
blood leukocytes following in vitro exposure to THz radiation:
A pilot study. Health Phys 2007;92(4):349–357.
42. Korenstein-Ilan A, Barbul A, Hasin P, Eliran A, Gover A,
Korenstein R. Terahertz radiation increases genomic insta-
bility in human lymphocytes. Radiat Res 2008; 224–234.
43. Berns MW, Bewley W. Inhibition of nucleic acid synthesis in
cells exposed to 200 micrometer radiation from the free
electron laser. Photochem Photobiol 1987;46(2):165–167.
44. Scarfi MR, Roman M, Di Pietro R, Zeni O, Doria A, Gallerano
GP, Giovenale E, Messina G, Lai A, Campurra G, Coniglio D,
D’Arienzo M. THz exposure of whole blood for the study of
biological effects on human lymphocytes. J Biol Phys
162WILMINK ET AL.
45. Homenko A, Kapilevich B, Kornstein R, Firer MA. Effects of Download full-text
100GHz radiation on alkaline phosphatase activity and
46. Frohlich H. The extraordinary dielectric properties of bio-
logical materials and the action of enzymes. Proc Natl Acad
Sci USA 1975;72(11):4211–4215.
47. Alexandrov BS, Gelev V, Bishop AR, Usheva A, Rasmussen
KO. DNA breathing dynamics in the presence of a terahertz
field. Biol Phys 2010;374(10):1214–1217.
48. Chitanvis SM. Can low-power electromagnetic radiation
disrupt hydrogen bonds in dsDNA? J Polym Sci B 2006;
49. Wilmink GJ, Opalenik SR, Beckham JT, Davidson JM,
Jansen ED. Assessing laser-tissue damage with biolumines-
cent imaging. J Biomed Optics 2006;11(4):041114.
50. Wilmink GJ, Opalenik SR, Beckham JT, Abraham AA,
Nanney LB, Mahadevan-Jansen A, Davidson JM, Jansen
ED. Molecular imaging-assisted optimization of hsp70
expression during laser-induced thermal preconditioning
for wound repair enhancement. J Invest Dermatol 2009;
51. Bouguer P. In: Jombert C, editor. Essai d’Optique sur la
gradation de la Lumiere. Paris, France: 1729; 180.
52. Ibey BL, Payne JA, Mixon DG, Thomas RJ, Roach WP.
Comparing finite difference time domain and Monte Carlo
modeling of human skin interaction with terahertz radiation.
In: Jacques SL, Roach WP, Thomas PJ, editors. Optical
interactions with tissue and cells XIX. Bellingham, WA:
Society of Photo-Optical Instrumentation Engineers; 2008.
53. Erdreich LS, Klauenberg BJ. Radio frequency radiation
exposure standards: Considerations
Health Phys 2001;80(5):430–439.
54. (ICNIRP), I.c.o.n.-i.r.p. Guidelines on limits of exposure to
laser radiation of wavelengths between 180nm and 1000mm.
Health Physics 1996;71(5):804–819.
55. Mantiply ED, Pohl KR, Poppell SW, Murphy JA. Summary of
measured radiofrequency electric and magnetic fields (10kHz
to 30GHz) in the general and work environment. Bioelec-
56. Protection, I.C.o.N.-I.R. ICNIRP statement on far infrared
radiation exposure. Health Phys 2006;91(6):630–645.
57. Union, T.E.P., a.t.C.o.t.E. Directive 2006/25/EC of the Euro-
pean Parliament and of the Council. Off J Eur Union 2006;
58. Morimoto RI. Cells in stress: Transcriptional activation of
heat shock genes. Science 1993;259(5100):1409–1410.
59. Morimoto RI. Dynamic remodeling of transcription com-
plexes by molecular chaperones. Cell 2002;110(3):281–284.
60. Morimoto RI. Proteotoxic stress and inducible chaperone
networks in neurodegenerative disease and aging. GenesDev
61. Morimoto RI, Kline MP, Bimston DN, Cotto JJ. The heat-
shock response: Regulation and function of heat-shock
proteins and molecular chaperones. Essays Biochem 1997;
62. Morimoto RI, Kroeger PE, Cotto JJ. The transcriptional
regulation of heat shock genes: A plethora of heat shock
factors and regulatory conditions. Experientia Suppl 1996;
63. Morimoto RI, Santoro MG. Stress-inducible responses and
heat shock proteins: New pharmacologic targets for cytopro-
tection. Nat Biotechnol 1998;16(9):833–838.
64. Stubbert LJ, Spronck JC, Hamill JD, McKay BC. Ultraviolet
light induces the sustained unscheduled expression of cyclin
E in the absence of functional p53. Cell Cycle 2009;8(18):
65. Hale GM, Querry MR. Optical constants of water in the 200-
nm to 200-microm wavelength region. Appl Opt 1973;12(3):
66. Ramundo-Orlando A, Gallerano GP, Stano P, Doria A,
Giovenale E, Messina G, Cappelli M, D’Arienzo M, Spassov-
sky I. Permeability changes induced by 130GHz pulsed
radiation on cationic liposomes loaded with carbonic anhy-
drase. Bioelectromagnetics 2007;28(8):587–598.
67. Alexandrov BS. Pre-melting dynamics of DNA and its
relation to specific functions. J Phys 2009;21(3):034107.
68. Clothier RH, Bourne N. Effects of THz exposure on human
primary keratinocyte differentiation and viability. J Biol
69. Olshevskaya JS, Ratushnyak AS, Petrov AK, Kozlov AS,
Zapara TA. Effect of terahertz electromagnetic waves on
neurons systems. Computational Technologies in Electrical
and Electronics Engineering, 2008. SIBIRCON 2008. IEEE
Region 8 International Conference on 2008: p. 210–211.
TERAHERTZ BIOEFFECTS IN DERMAL FIBROBLASTS163