Advertisement

Topology-Based Prediction of Pathway Dysregulation Induced by Intense Terahertz Pulses in Human Skin Tissue Models

  • Cameron M. Hough
  • David N. Purschke
  • Chenxi Huang
  • Lyubov V. Titova
  • Olga Kovalchuk
  • Brad J. Warkentin
  • Frank A. Hegmann
Article
  • 50 Downloads

Abstract

The strong interaction between terahertz (THz) radiation and biological systems has motivated the development of several biomedical technologies, including imaging and spectroscopy applications with promising potential for improved disease diagnosis. This interaction mechanism also implies that external excitation with intense pulses of THz energy could couple to important biological structures and induce significant downstream phenotypic effects. In this study, we expose human skin tissue models to a prolonged train of high-intensity THz pulses and measure the resulting global differential gene expression. From these data, signal pathway perturbation analysis identified pathways that are predicted to be significantly dysregulated, including the cytokine-cytokine receptor interaction and glioma pathways, and further identified the gene-level mechanisms predominantly responsible. These results indicate that induction of an inflammatory-like response and suppression of division/differentiation in cancer are possible. These effects could be further explored and characterized in different types of normal and cancerous tissues to determine potential novel clinical applicability of intense THz pulses.

Keywords

Terahertz Intense terahertz pulses Global gene expression Skin Bioinformatics Pathway dysregulation Inflammatory response Glioma Cancer 

1 Introduction

Terahertz (THz) frequencies strongly couple to natural low-frequency oscillations of many important biomolecules, and the stretching/twisting modes of hydrogen bond networks that are ubiquitous in biological systems (e.g., water, proteins, and DNA) [1]. This interaction mechanism makes THz a highly sensitive probe of molecular structure in biological systems, which has led to the development of diagnostic imaging/spectroscopy technologies with excellent contrast between healthy and diseased tissue resulting in high sensitivity/specificity in diagnosing several human cancers [2, 3, 4, 5, 6, 7]. However, several studies have also shown that THz interactions can induce significant biological changes [8, 9]. Echchgadda et al. have investigated thermal effects induced in vitro for varying CW THz frequencies and observed frequency-dependent genomic expression responses nearly entirely distinct when compared to equivalent uniform bulk heating in human keratinocytes [10, 11]. Studies investigating non-thermal effects have found that THz pulses alter gene expression at the transcript and protein level in human skin tissue models [12], increase membrane permeation [13], affect cellular differentiation in mouse stem cells [14], or induce severe forms of genotoxic stress at the DNA [15] or chromosome [16] level in skin cells. These effects may indicate the potential for currently unknown health risks, for which safe exposure levels should be determined, or lead to phenotypic effects that could be exploited for novel diagnostic/therapeutic clinical application.

In this paper, we outline the technical details and results of an investigation of the dysregulation in gene networks and signaling pathways in human skin models induced by prolonged exposure to a train of intense THz pulses. Illumina Whole Genome BeadChip arrays were used to analyze global gene expression changes at the transcript (mRNA) level in THz-exposed samples. Since mRNA is a precursor to a functional protein, biological effects are predicted based on the known interaction characteristics of the corresponding proteins. Bioinformatic analyses allow for prediction of the possible phenotypic endpoint by calculation of the magnitude and direction (i.e., activation vs. inhibition) of the expected perturbation in known gene interaction networks. The pathways corresponding to the greatest amount of predicted dysregulation are identified and discussed.

2 Experimental Methods

2.1 Generation and Detection of Intense THz Pulses

Several methods of generating high-intensity THz pulses are available with varying capabilities depending on the experimental design [17]: plasma sources are capable of producing intense broadband pulses at high THz frequencies (< 30 THz), which allow for shorter, more tightly focused pulses and greater coupling efficiency to faster biological dynamics. However, the low spectral density results in weak coupling to a wider range of modes, and the small spot size exposes a proportionately reduced cell population. Organic crystals (e.g., DAST) are capable of achieving very large THz fields at intermediate frequencies (~ 1–10 THz), but the limited repetition rate (< 10–100 Hz) requires extended exposure times for equivalent effect.

In this study, intense THz pulses are generated by optical rectification in lithium niobate (LiNbO3) due to the capability of high repetition rates (> 1 kHz) and large spectral density at low THz frequencies that strongly interact with dense hydrogen bond networks and natural oscillations of large macromolecules. Infrared pulses (800 nm, 50 fs, 3.6 mJ) are generated and amplified at a repetition rate of 1 kHz by a Ti:sapphire oscillator and regenerative amplifier (Coherent Micra/Legend). The beam is partitioned such that 80% of the energy is utilized for THz generation, while a small amount of the remaining power is used as the probe for THz detection via electro-optic sampling (discussed next). The generation line is directed towards an 1800-ln/mm reflective diffraction grating, as shown in Fig. 1, causing angular dispersion (i.e., the “tilted-pulse-front”) [18, 19], and a telescopic 4f-lens system images both the pulse front and the grating plane onto the face of a LiNbO3 crystal with a 63° cut angle. Optical rectification of the 800-nm laser pulse in the LiNbO3 crystal induces a time-varying nonlinear polarization that produces a single-cycle pulse that corresponds to a broad band of THz frequencies [20]. Following THz pulse generation, a layer of black polyethylene blocks any remaining visible/infrared wavelengths but allows the THz pulse to pass through. Two-inch-(50.8 mm)-diameter off-axis gold parabolic mirrors collect and focus the THz energy to the smallest achievable spot size (highest intensity) to either the sample position or the detection crystal face. The nominal f-number of the focusing mirror is 1.5.
Fig. 1

Photograph of the THz generation scheme used to generate intense THz pulses in the Ultrafast Nanotools Lab at the University of Alberta

The temporal profile of the THz waveform is measured by electro-optic sampling in a 0.2-mm-thick (110) gallium phosphide (GaP) crystal on a 2 mm (100) GaP substrate. The electro-optic modulation ΔI/2I0 is measured for varying delay times with temporal resolution of 0.1 ps (fmax= 5 THz) and a total duration of 90 ps (Δf = 6 GHz). With knowledge of the GaP refractive index (n0= 3.18), GaP transmission coefficient (tGaP= 0.46), electro-optic coefficient (r41= 0.88 pm/V), and crystal thickness (L), the electric field ETHz is calculated from [21].
$$ \varDelta I/2{I}_0=\sin \left(2\pi {n}_0^3{r}_{41}{t}_{\mathrm{GaP}}{E}_{\mathrm{THz}}L/{\lambda}_0\right) $$
(1)
The total pulse energy is measured using a pyroelectric detector (SPJ-D-8, Spectrum Detector, Inc.), and the transverse intensity distribution of the THz focus is mapped with a pyroelectric camera (PV320, Electrophysics). The pulse train, waveform, frequency spectrum, and spot image characterized using these methods are shown in Fig. 2. The THz pulse energy and peak electric field was 2.4 μJ and 240 kV/cm, respectively, corresponding to power densities of 73.8 MW/cm2 (peak) and 74 mW/cm2 (average). The temperature increase in water from a single THz pulse is estimated to be ~ 0.4 × 10−3 °C, and the average heating induced by the 1-kHz train of THz pulses was measured using a thermal imaging camera to be < 1 °C, as is typical for similar studies [8, 12]. Additionally, no biological activity (gene expression, gene ontology association, signaling pathway dysregulation) related to heat response or thermal shock was identified. Due to the relatively low temperature increase and lack of heating response, the observed effects reported are therefore not thought to be thermally induced.
Fig. 2

(a) Diagram of the pulse train, indicating the approximate pulse duration and repetition time. (b) Electric field vs. time for each pulse. The pulse energy is 2.4 μJ. Inset: Transverse intensity distribution (1/e2 spot size = 1.6 × 2.7 mm2). (c) Corresponding frequency spectrum. The peak frequency is 0.6 THz and the FWTM bandwidth is 1.6 THz

2.2 Sample and Exposure Details

3D human skin models were obtained from the MatTek corporation (EpidermFT, www.mattek.com). These are full-thickness 3D artificial models of human skin, derived from co-cultures of human-derived epidermal keratinocytes and dermal fibroblasts in a collagen matrix. The stratified growth and the metabolic/mitotic properties closely model the biological activity of human skin in vivo [22]. Tissues were stored and handled according to the manufacturer’s instructions. Briefly, upon delivery, samples were transferred to wells containing fresh warm media such that the top surface was exposed to air, and the bottom surface absorbed media through a thin porous membrane. All wells were placed in an incubator overnight to equilibrate. For exposures, a sample holder was centered over the THz focus by maximizing energy/field through an attached pinhole iris, and samples were positioned such that the THz beam focused to the epidermal layer. Each sample was individually exposed for 10 min (44.4 J/cm2 total accumulated energy density). Following THz exposure, samples were returned to wells with fresh media, incubated (37 °C, 5% CO2) for 30 min, and fixed by snap-freezing in liquid nitrogen. Sham-exposed controls went through the exact same process, but with the THz beam fully blocked. Each set of exposure parameters was repeated in quadruplicate.

2.3 Measurement of Global Differential Gene Expression

The central exposed region (1.5-mm diameter) of the frozen tissue samples was excised with a sterile punch tool aligned using the same pinhole iris used for exposure alignment, and total RNA was isolated as per the manufacturers’ instructions (Zymo Research, Irvine, CA). Following elution in DNase/RNase-free ultrapure water, the quality was quantified with UV spectroscopy (NanoDrop, Wilmington, DE) and assessed for RNA integrity (Agilent 2100 Bioanalyzer, Santa Clara, CA). cDNA was amplified (Ovation FFPE WTA), labeled (Encore BiotinIL), and hybridized to the HumanHT-12_v4 Whole Genome Expression BeadChip arrays according to the manufacturer’s instructions (Illumina, San Diego, CA). BeadChips were imaged, the fluorescence signal for each probe was quantified using the iScan platform (Illumina), and data were normalized/analyzed using the Illumina BeadStudio software. Differential expression (fold-change) was quantified, and corresponding p values were corrected for multiple hypothesis testing via the false discovery rate (FDR) method. Data were corrected for batch effects via the ComBat algorithm (http://www.bu.edu/jlab/wp-assets/ComBat/Abstract.html).

2.4 Topology-Based Signal Pathway Perturbation Analysis

Signal pathway impact analysis (SPIA) was applied using the R/Bioconductor bioinformatics pathway analysis package ROntoTools, as described in references [23, 24, 25] and outlined in Fig. 3. This method estimates how the measured changes in gene expression will dysregulate biological processes by propagating the total perturbation through known pathways, accounting for gene-level interactions within the considered network. Gene interaction networks for these calculations were acquired from the KEGG database (reference [26], http://www.genome.jp/kegg/pathway.html), with a focus on pathways that are in some way related to the initiation, maintenance, or progression of human cancer. For a given network, a perturbation factor
$$ \mathrm{PF}\left({g}_i\right)=\varDelta E\left({g}_i\right)+A\left({g}_i\right) $$
(2)
is calculated for each gene gi, where ΔE(gi) is the measured log2-fold expression change from the microarray data. A(gi) represents the accumulated perturbation due to upstream perturbations in the signaling network, estimated as
$$ A\left({g}_i\right)={\sum}_j{\beta}_{ij}\times PF\left({g}_j\right)/{N}_{\mathrm{ds}} $$
(3)
where βij is the interaction matrix describing the relationship between the ith and jth gene, and Nds is the number of genes downstream of gj. The total perturbation status for each pathway is calculated from Eq. (3) as
$$ {A}_{\mathrm{tot}}={\sum}_iA\left({g}_i\right) $$
(4)
Fig. 3

Sample workflow for the calculation performed by ROntoTools described in Section 2. (a) Measured global differential gene expression induced by intense THz pulses (see Fig. 4); (b) the pathway information from the KEGG database is used to filter the data in (a) to only include genes in the relevant network; (c) pathway topology information (node/edge interaction properties) is used to calculate the upstream accumulated perturbation A(gi) for each gene in the network and is presented as a two-way plot (measured gene expression vs. calculated accumulated perturbation, see Fig. 5); (d) the total accumulated perturbation is determined (blue line) and compared to the simulated null distribution (black curve). If the accumulated perturbation is greater than the quantity corresponding to the pre-defined significance threshold determined from the null distribution (yellow dot), the calculated perturbation is considered statistically significant, and the pathway is predicted to be dysregulated by the relative magnitude Atot (Eq. 4)

Corresponding p values are determined via a bootstrapping simulation as described in [24] and adjusted for multiple hypothesis testing with the FDR method. Significantly dysregulated pathways for which Atot is positive are predicted to be “activated,” and pathways for which Atot is negative are predicted to be “suppressed.” This procedure allows for prediction of the relative magnitudes by which pathways are affected, the genes most sensitive to THz exposure, and the genes that are predominantly responsible for driving the predicted pathway dysregulation.

3 Results

3.1 Global Differential Gene Expression Induced by Intense THz Pulses

Figure 4 shows the global differential gene expression induced by 10-min exposure to intense THz pulses. Of 9311 total genes detected in control probes, 593 were found to be upregulated (red) and 1088 were found to be downregulated.
Fig. 4

Volcano plot displaying the measured global differential gene expression profile induced by intense THz pulses in human skin tissue models. The statistical significance of each measurement is plotted as the adjusted p value (negative log10-transformed) against the magnitude of differential gene expression (log2-transformed fold-change of exposed vs. control tissues). Dotted black lines indicate conventional criteria for significance of gene expression. Of 9311 total genes detected in control probes, 1088 were downregulated and 593 were upregulated

These data were used in combination with information from the KEGG database of pathway network information to calculate the predicted perturbative effects in 184 distinct gene interaction networks as described in Section 2.

3.2 Pathway Perturbation Analysis

The pathways predicted to be most significantly dysregulated by intense THz pulses are (1) cytokine-cytokine receptor interaction pathway and (2) glioma pathway.

The total accumulated perturbation calculated for the cytokine-cytokine receptor interaction pathway was Atot =  + 23.3 (p = 0.004) and so is predicted to be activated by exposure to intense THz pulses. Cytokines are secreted by cells in response to some stimuli, usually as an innate or adaptive inflammatory defense [27]. The interaction with cytokine receptors activates cellular processes involved in inflammatory responses to external stimuli.

The total accumulated perturbation calculated for the glioma pathway was Atot =  − 36.8 (p = 0.004) and is therefore predicted to be suppressed by exposure to intense THz pulses. This pathway describes the gene-level mechanisms involved in initiation, progression, and metastatic activity of glioma, a type of brain cancer affecting glial cells [27]. As will be discussed, although the glioma pathway describes molecular signaling in neural tissue cell types, the genes that are affected are ubiquitous in a wide variety of eukaryotic cell/tissue types, including skin.

4 Discussion

Several studies have now been performed that quantify global differential gene expression using similar significance thresholds for varying THz pulse train parameters. Estimates of the peak/average intensity from these publications show that the genomic expression response to THz pulses tends to increase roughly with the pulse intensity, as shown in Table 1. It is noted that these studies investigated different biological systems and utilized different THz exposure conditions, and as such should only be approximately compared for magnitude of effect.
Table 1

Results of global differential gene expression analysis for varying THz intensities in the literature

Peak intensity (MW/cm2)

Average intensity (mW/cm2)

Number of genes differentially expressed

Reference

73.8

74

1681

This study

33

57

442

[12]

2

~0

20

[28]

1a

~0

149

[29]

0.002

0.12

0

[30]

aPeak intensity calculated from the average power, pulse duration, and repetition rate reported in reference [29]

Figure 5 shows the gene expression and resulting gene-level perturbation for the highest significance pathways dysregulated by intense THz pulses. For the cytokine-cytokine receptor interaction pathway (Fig. 5(a), http://www.genome.jp/kegg-bin/show_pathway?hsa04060 [27]), significant contribution to the total accumulated perturbation (Atot) was due to genes that were not directly affected by the THz exposures, but rather were downstream targets of directly affected genes. This is due to the genes directly affected encoding for cytokines that bind to downstream cell receptor targets, such as:
  • chemokines (CXCL family, CCL family)
    • Released in several cell types undergoing inflammatory stimuli

    • Involved in mediating inflammatory responses by guiding neutrophils to target site

  • interleukins (IL6)
    • Wide variety of biological functions

    • Primarily produced at acute inflammation sites and induces inflammatory response (required for generation of helper T cells)

  • interferons (IL24)
    • Over-expression correlates with apoptosis induction

    • Anti-proliferative properties in melanoma cells

Fig. 5

Two-way plots showing measured differential gene expression vs. calculated accumulated perturbation for genes involved in (a) cytokine-cytokine receptor interaction and (b) glioma pathways. Green points correspond to genes that were not affected by THz but were affected by upstream perturbations in the relevant pathway. Blue points are genes that were differentially expressed by THz exposures and therefore contributed to downstream perturbation, but were not perturbed by upstream perturbations. Red points are genes that were both differentially expressed by THz exposures and were further affected by upstream perturbations in the associated gene interaction network

The upregulation of these genes and predicted activation of the corresponding pathway indicate that prolonged exposure to intense THz pulses may induce an inflammatory-like response in human skin.

In the glioma pathway (Fig. 5(b), http://www.genome.jp/kegg-bin/show_pathway?hsa05214 [27]), the majority of the predicted suppression is due to genes that were also directly affected by THz exposures. These include the following:
  • calmodulin family (CALM/CALML, CAMK subfamilies)
    • Encode for calcium-binding messenger proteins expressed in eukaryotic cells

    • Active in signal transduction pathways regulating a diverse set of biological functions, including regulating cell growth/proliferation in the glioma pathway

  • KRAS and ERK
    • Encode for proteins in the canonical mitogen-activated protein kinase (MAPK) pathway, which regulates proliferation/differentiation and is one of the most commonly implicated pathways in human cancer [31].

These results report gene-/pathway-level information that regulate the effects observed in reference [12], in which a downregulation of genes was predicted to largely suppress epidermal differentiation. It is recognized that glioma (and the initiation/progression regulated by the glioma pathway) is a cancer affecting cell types corresponding to neural tissue, whereas these predictions are based on observations in skin tissue models. However, the gene-level mechanisms responsible for the negative perturbation are genes that encode for proteins involved in calcium and MAPK signaling, and these are ubiquitous and well-conserved across many different cell types including skin and neural cells [32, 33]. The downregulation of these genes and predicted suppression of the glioma pathway indicate that THz exposures may play a role in suppression of division, differentiation, and progression of this cancer if this effect is found to occur in the relevant cell type.

To pursue these results, future studies should focus on an investigation of these potential effects in glial cells (or other neural tissue cells), for both normal and cancerous phenotypes, which may elucidate presently unknown clinical applications of intense THz pulses. Furthermore, these results were obtained at a relatively early time-point (30-min following exposure), and so a characterization of later time-points (several hours to days following exposure) probing prolonged effects on division and differentiation would be of significant interest.

5 Conclusion

In this paper, we show that high-intensity pulses of THz radiation induce a large genomic expression response in human skin tissue models, and this effect tends to logically scale when compared to THz intensities utilized in similar studies (Table 1). Furthermore, pathway perturbation analysis using recently developed bioinformatics analysis tools allows for the phenotypic endpoint to be predicted from the profile of differentially expressed genes [23, 24, 25]. The pathways corresponding to the largest significant dysregulation were identified in order to predict the dominant effect of prolonged exposure to intense THz pulses. In the considered pathways, the largest positive perturbation (i.e., activation) was observed in the cytokine-cytokine receptor interaction pathway, indicating that an acute inflammatory response may be induced by the strong external stimulus of the very high peak electric field. The pathway corresponding to the largest significant negative perturbation (i.e., inhibition) was the glioma pathway, indicating a potential suppression of growth, division, and differentiation in molecular signaling pathways that drive glioma progression. The genes measured to be predominantly responsible for these effects were identified, and the observations are consistent with results of studies utilizing similar experimental designs [12, 29]. A relevant pursuit of these findings involves investigation of the effect of intense THz pulses on division/differentiation dynamics in neural cell types. If suppression of growth is observed in these (or other) systems, especially in cancerous tissues, intense THz pulses may potentially find novel clinical application with the goal of targeted inhibition of mitotic activity in cancerous cells.

Notes

Acknowledgements

We acknowledge support from NSERC, CFI, and the AITF Strategic Chairs Program, and technical assistance from Beipei Shi, Greg Popowich, Matt Reid, Rocio Rodriguez-Juarez, Rommy Rodriguez-Juarez, Andrey Golubov, and Yaroslav Ilnytskyy.

References

  1. 1.
    Son, J.-H. (2014). Terahertz Biomedical Science & Technology. Boca Raton, FL: CRC Press. Print.CrossRefGoogle Scholar
  2. 2.
    Cheon, H., Yang, H.-J., & Son, J.-H. Toward Clinical Cancer Imaging Using Terahertz Spectroscopy. IEEE Journal of Selected Topics in Quantum Electronics. 23(4): 2017.Google Scholar
  3. 3.
    Ashworth, P.C., Pickwell-MacPherson, E., Provenzano, E., Pinder, S.E., Purushotham, A.D., Pepper, M., & Wallace, V.P. Terahertz pulsed spectroscopy of freshly excised human breast cancer. Optics Express. 17(15): 12444–12454, 2009.Google Scholar
  4. 4.
    Wallace, V.P., Fitzgerald, A.J., Shankar, S., Flanagan, N. Pye, R., Cluff, J., & Arnone, D.D. Terahertz pulsed imaging of basal cell carcinoma ex vivo and in vivo. British Journal of Dermatology. 151(2): 424–432, 2004.CrossRefGoogle Scholar
  5. 5.
    Pickwell-MacPherson, E. & Wallace, V.P. Terahertz pulsed imaging – A potential medical imaging modality? Photodiagnosis and Photodynamic Therapy. 6(2): 128–134, 2009.CrossRefGoogle Scholar
  6. 6.
    Teraview Application Note. Using the TPS Spectra 3000 as a Tool for the Investigation of Cancers: Using the Hand-held Probe to Investigate Breast Cancer. 2012.Google Scholar
  7. 7.
    Eadie, L., Reid, C., Fitzgerald, A., & Wallace, V. Optimizing multi-dimensional terahertz imaging analysis for colon cancer diagnosis. Expert Systems with Applications. 40(6): 2043–2050, 2013.CrossRefGoogle Scholar
  8. 8.
    Wilmink, G.J. & Grundt, J. E. Invited Review Article: Current State of Research on Biological Effects of Terahertz Radiation. Journal of Infrared, Millimeter, and Terahertz Waves. 32: 1074–1122, 2011.CrossRefGoogle Scholar
  9. 9.
    Romanenko, S., Begley, R., Harvey, A., Hool, L., & Wallace, V. The interaction between electromagnetic fields at megahertz, gigahertz, and terahertz frequencies with cells, tissues, and organisms: risks and potential. Journal of Royal Society Interface. 14: 20170585.  https://doi.org/10.1098/rsif.2017.0585
  10. 10.
    Echchgadda, I., Grundt, J.E., Cerna, C.Z., Roth, C.C., Payne, J.A., Ibey, B.L., & Wilmink, G.J. Terahertz Radiation: A Non-contact Tool for the Selective Stimulation of Biological Responses in Human Cells. IEEE Transactions on Terahertz Science and Technology. 6(1): 54–68, 2016.CrossRefGoogle Scholar
  11. 11.
    Echchgadda, I., Cerna, C.Z., Sloan, M.A., Elam, D.P., & Ibey, B.L. Effects of Different Terahertz Frequencies on Gene Expression in Human Keratinocytes. Proc. Of SPIE. Optical Interactions with Tissue and Cells XXVI. 9321: 1–9, 2015.Google Scholar
  12. 12.
    Titova, L.V., Ayesheshim, A.K., Golubov, A., Rodriguez-Juarea, R., Woycicki, R., Hegmann, F.A., & Kovalchuk, O. Intense THz pulses down-regulate genes associated with skin cancer and psoriasis: a new therapeutic avenue? Scientific Reports. 3(2362): 1–6, 2013.Google Scholar
  13. 13.
    Gallerano, G.P. et al. THz-BRIDGE: Final Report. Quality of Life Management/Living Resources. 2004.Google Scholar
  14. 14.
    Bock, J., Fukuyo, Y., Kang, S., Phipps, M.L., Alexandrov, L.B., Rasmussen, K.O., Bishop, A.R., Rosen, E.D., Martinez, J.S., Chen, H.-T., Rodriguez, G., Alexandrov, B.S., Usheva, A. Mammalian Stem Cells Reprogramming in Response to Terahertz Radiation. PLoS One. 5(12): 1–6, 2010.CrossRefGoogle Scholar
  15. 15.
    Titova, L.V., Ayesheshim, A.K., Golubov, A., Fogen, D., Rodriguez-Juarea, R., Hegmann, F.A., & Kovalchuk, O. Intense THz pulses cause H2AX phosphorylation and activate DNA damage response in human skin tissue. Biomedical Optics Express. 4(4): 559–568, 2013.CrossRefGoogle Scholar
  16. 16.
    Amicis, A., De Sanctis, S., Di Cristofaro, S., Franchini, V., Lista, F., Regalbuto, E., Giovenale, E., Gallerano, G.P., Nenzi, P., Bei, R., Fantini, M., Benvenuto, M., Masuelli, L., Coluzzi, E., Cicia, C., & Sgura, A. Biological effects of in vitro THz radiation exposure in human foetal fibroblasts. Mutation Research/Genetic Toxicology and Environmental Mutagenesis. 793: 150–160, 2015.CrossRefGoogle Scholar
  17. 17.
    Hoffman, M.C. “Intense laser-based THz sources” in The 2017 Terahertz Science and Technology Roadmap. Journal of Applied Physics D. 50: 6–7, 2017.Google Scholar
  18. 18.
    Hebling, J., Almasi, G., & Kozma, I.Z. Velocity matching by pulse front tilting for large-area THz-pulse generation. Optics Express. 10(21): 1161–1166, 2002.CrossRefGoogle Scholar
  19. 19.
    Yeh, K.-L., Hoffman, M.C., Hebling, J., & Nelson, K.A. Generation of 10 μJ ultrashort terahertz pulses by optical rectification. Applied Physics Letters. 90(171121): 1–3, 2007.Google Scholar
  20. 20.
    Lee, Y. (2009). Principles of Terahertz Science and Technology. New York, NY: Springer. Print.Google Scholar
  21. 21.
    Hirori, H., Doi, A., Blanchard, F., & Tanaka, K. Single-cycle terahertz pulses with amplitudes exceeding 1 MV/cm generated by optical rectification in LiNbO3. Applied Physics Letters. 98(9): 1–3, 2011.Google Scholar
  22. 22.
    Acheva, A., Aerts, A., Rombouts, Ch., Baatout, S., Salomaa, S., Manda, K., Hildebrandt, G., & Kamarainen, M. Human 3-D tissue models in radiation biology: current status and future perspectives. International Journal of Radiation Research. 12(2): 81–98, 2014.Google Scholar
  23. 23.
    Voichita C, Ansari S and Draghici S (2017). ROntoTools: R Onto-Tools suite. R package version 2.6.0.Google Scholar
  24. 24.
    Tarca, A.L., Draghici, S., Khatri, P., Hassan, S.S., Mittal, P., Kim, J.-s., Kim, C.J., Kusanovic, J.P., & Romero, R. A novel signaling pathway impact analysis. Bioinformatics. 25(1): 75–82, 2009.CrossRefGoogle Scholar
  25. 25.
    Voichita, C., Donato, M., & Draghici, S. Incorporating gene significance in the impact analysis of signaling pathways. 11 th International Conference on Machine Learning and Applications. ICMLA 2012: 126–131, 2012.Google Scholar
  26. 26.
    Kanehisa, M. & Goto, S. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Research. 28(1): 29–34, 2000.CrossRefGoogle Scholar
  27. 27.
    Website: http://www.genome.jp/kegg/kegg1.html [Electronic access]
  28. 28.
    Alexandrov, B. et al. Specificity and Heterogeneity of Terahertz Radiation Effect on Gene Expression in Mouse Mesenchymal Stem Cells. Scientific Reports. 3(1184): 1–8, 2013.Google Scholar
  29. 29.
    Kim, K.-T., Park, J., Jo, S.J., Jung, S., Kwon, O.S., Gallerano, G.P., Park, W.-Y., & Park, G.-S. High-power femtosecond-terahertz pulse induces a wound response in mouse skin. Scientific Reports. 3(2296): 1–7, 2013.Google Scholar
  30. 30.
    Williams, R. The influence of high intensity terahertz radiation on mammalian cell adhesion, proliferation, and differentiation. Physics in Medicine and Biology. 58(2): 373–391, 2013.CrossRefGoogle Scholar
  31. 31.
    Takashima, A. & Faller, D. Targeting the RAS Oncogene. Expert Opinion on Therapeutic Targets. 17(5): 507–31. 2013.CrossRefGoogle Scholar
  32. 32.
    Fecher, L.A., Amaravadi, R.K., & Flaherty, K.T. The MAPK pathway in melanoma. Current Opinion in Oncology. 20(2): 183–189, 2008.CrossRefGoogle Scholar
  33. 33.
    Pandey, V., Bhaskara, V.K., & Babu, P.P. Implications of Mitogen-Activated Protein Kinase Signaling in Glioma. Journal of Neuroscience Research. 94(2): 114–127, 2016.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Medical Physics Division, Department of OncologyUniversity of AlbertaEdmontonCanada
  2. 2.Department of PhysicsUniversity of AlbertaEdmontonCanada
  3. 3.Department of PhysicsWorcester Polytechnic InstituteWorcesterUSA
  4. 4.Department of BiologyUniversity of LethbridgeLethbridgeCanada

Personalised recommendations