biochemical signatures of radiation response in tumour ...agbrolo/pmb_quinn_2011_nov.pdf ·...

Biochemical signatures of in vitro radiation response in human lung breast and prostate

tumour cells observed with Raman spectroscopy

This article has been downloaded from IOPscience Please scroll down to see the full text article

2011 Phys Med Biol 56 6839

(httpiopscienceioporg0031-91555621006)

Download details

IP Address 14210411558

The article was downloaded on 09112011 at 2111

Please note that terms and conditions apply

View the table of contents for this issue or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING PHYSICS IN MEDICINE AND BIOLOGY

Phys Med Biol 56 (2011) 6839ndash6855 doi1010880031-91555621006

Biochemical signatures of in vitro radiation responsein human lung breast and prostate tumour cellsobserved with Raman spectroscopy

Q Matthews1 A Jirasek1 J J Lum2 and A G Brolo3

1 Department of Physics and Astronomy University of Victoria Victoria BC V8W 3P6 Canada2 Deeley Research Centre BC Cancer Agency Vancouver Island Centre Victoria BC V8R 6V5Canada3 Department of Chemistry University of Victoria Victoria BC V8W 3V6 Canada

E-mail qmatthewuvicca and jirasekuvicca

Received 7 July 2011 in final form 5 September 2011Published 5 October 2011Online at stacksioporgPMB566839

AbstractThis work applies noninvasive single-cell Raman spectroscopy (RS) andprincipal component analysis (PCA) to analyze and correlate radiation-inducedbiochemical changes in a panel of human tumour cell lines that vary by tissueof origin p53 status and intrinsic radiosensitivity Six human tumour cell linesderived from prostate (DU145 PC3 and LNCaP) breast (MDA-MB-231 andMCF7) and lung (H460) were irradiated in vitro with single fractions (15 30 or50 Gy) of 6 MV photons Remaining live cells were harvested for RS analysisat 0 24 48 and 72 h post-irradiation along with unirradiated controls Single-cell Raman spectra were acquired from 20 cells per sample utilizing a 785 nmexcitation laser All spectra (200 per cell line) were individually post-processedusing established methods and the total data set for each cell line was analyzedwith PCA using standard algorithms One radiation-induced PCA componentwas detected for each cell line by identification of statistically significantchanges in the PCA score distributions for irradiated samples as compared tounirradiated samples in the first 24ndash72 h post-irradiation These RS responsesignatures arise from radiation-induced changes in cellular concentrations ofaromatic amino acids conformational protein structures and certain nucleicacid and lipid functional groups Correlation analysis between the radiation-induced PCA components separates the cell lines into three distinct RS responsecategories R1 (H460 and MCF7) R2 (MDA-MB-231 and PC3) and R3(DU145 and LNCaP) These RS categories partially segregate according toradiosensitivity as the R1 and R2 cell lines are radioresistant (SF2 gt 06) andthe R3 cell lines are radiosensitive (SF2 lt 05) The R1 and R2 cell lines furthersegregate according to p53 gene status corroborated by cell cycle analysis post-irradiation Potential radiation-induced biochemical response mechanismsunderlying our RS observations are proposed such as (1) the regulated synthesis

0031-915511216839+17$3300 copy 2011 Institute of Physics and Engineering in Medicine Printed in the UK 6839

6840 Q Matthews et al

and degradation of structured proteins and (2) the expression of anti-apoptosisfactors or other survival signals This study demonstrates the utility of RSfor noninvasive radiobiological analysis of tumour cell radiation response andindicates the potential for future RS studies designed to investigate monitor orpredict radiation response

S Online supplementary data available from stacksioporgPMB566839mmedia

1 Introduction

Optimizing the effectiveness of radiation therapy is limited in part by the variability in radiationresponse between patients Probabilities of both normal tissue complication and tumourcontrol depend on individual patient responses to treatment (Peters 1996) There is currentlyno proven method for assessing tumour radiation response in a patient during the course ofan extended treatment Efforts to develop a predictive assay for tumour radiation responseusing pretreatment indicators related to apoptosis (Levine et al 1995) intrinsic radiosensitivity(Levine et al 1995 West et al 1997 Bjork-Eriksson et al 2000) hypoxia (Nordsmark andOvergaard 2000 Vaupel and Mayer 2007 Luukkaa et al 2009) or tumour proliferation (Begget al 1999) have shown promise but most have had either unsatisfactory levels of success orposed significant technical difficulties preventing clinical implementation The relationshipbetween the genetic status of tumours and intrinsic radiosensitivity has been studiedextensively yet the importance of the functional state of certain genes (eg p53) in determiningradiosensitivity appears to depend on the tumour cell lines examined (McIlwrath et al 1994Oya et al 2003 Scott et al 2003 Williams et al 2007 2008) Recently developed methodsapplying multigene expression profiles to predict the tumour radiosensitivity of a patient bycomparisons with clonogenic survival data from established cell lines (Torres-Roca et al 2005Eschrich et al 2009) likely have the most potential for clinical implementation Such methodshave been shown to be statistically predictive of tumour response in esophageal and rectalcancers and of locoregional control in head and neck cancers (Eschrich et al 2009) Howeverboth pre-clinical and clinical studies report many false positives and negatives Such methodsmay also be inherently limited by the use of laboratory data from established tumour cell linesupon which the models are constructed possibly limiting the application for clinical casesacross a variety of tumour types In light of these previous and ongoing research efforts futureadvances in the field of experimental radiobiology as applied to personalized radiation therapymay benefit from the use of new biochemical analysis methods with the ability to analyzebiochemical radiation response in vitro or in vivo across a wide variety of biomolecules Onesuch technique is Raman spectroscopy (RS)

RS is a vibrational spectroscopy technique in which an optical wavelength laser isfocused onto a sample inducing transitions between molecular vibrational levels andcreating inelastically scattered photons (Raman scattering) with frequencies and intensitiescharacteristic of the molecules in the sample The resulting Raman spectrum provides adetailed description of the molecular composition within the sampling volume Within aspecific range of laser wavelength and power RS is noninvasive and nondestructive allowingthe analysis of live cells or tissues without perturbation of the sample (Puppels et al 1991Notingher et al 2002 2003) The use of high-power focusing optics can provide spatialresolutions as low as 1 μm well below the typical size of a human cell (10ndash50 μm diameter)Furthermore RS can provide biochemical information from unfixed cells at comparable levels

Biochemical signatures of radiation response in tumour cells observed with RS 6841

of accuracy and sensitivity as established techniques such as magnetic resonance spectroscopy(MRS) and flow cytometry (Mourant et al 2006) RS has an inherent advantage over otherbiochemical analysis tools by enabling the simultaneous detection of a variety of molecularstructures across proteins (eg amino acids conformational structures) nucleic acids (DNAand RNA) and lipids (eg cholesterol choline CH2 groups) in a single acquisition Thisutility allows complex biochemical changes in cells to be analyzed simultaneously acrossdifferent classes of biomolecules rather than analyzing a single type of molecule (eg DNA)or class of molecules (eg metabolites) an inherent restriction of many established techniquesRS also has the potential to be applied in vivo with the use of fiber-optic probe technology(Motz et al 2004) Two major drawbacks of RS for biological analysis are the inherentlyweak intensity of biological Raman scattering often necessitating long integration times andthe competition with fluorescence from biological materials optical components and samplesubstrates or containers However the use of high-power excitation lasers low-fluorescenceoptics and substrates and automated baseline removal methods (Lieber and Mahadevan-Jansen2003 Schulze et al 2005) can alleviate these difficulties

RS has been investigated in oncology for cancer detection and diagnosis successfullydiscriminating between healthy and cancerous skin (Nijssen et al 2002 Choi et al 2005Lieber et al 2008) bladder (de Jong et al 2006) and gastric (Teh et al 2008 2009) tissuesand aiding in the histopathological analysis of prostate cancer (Tollefson et al 2010) Otherstudies have applied RS for cellular biochemical analysis of apoptosis (Verrier et al 2004Zoladek et al 2011) necrosis (Kunapareddy et al 2008) cell death (Notingher et al 2003)non-proliferation (Short et al 2005) and cell cycle (Swain et al 2008 Matthews et al 2010)In radiobiological applications RS has detected molecular alterations in irradiated aqueousDNA (Sailer et al 1996 Shaw and Jirasek 2009) sodium hyaluronate (Synytsya et al 2011)biological membranes (Verma 1986 Verma and Rastogi 1990 Verma and Sonwalkar 1991Verma et al 1993 Sailer et al 1997) and skin and muscle tissues (Lakshmi et al 2002 Synytsyaet al 2004) A recent clinical RS study discriminated between responding and non-respondingcervical cancers post-irradiation (Vidyasagar et al 2008) yet made no conclusions regardingthe differences in the biochemical composition between tissues or the molecular basis forRS discrimination However the study does validate the potential for using RS in clinicalradiotherapy practice

Our recent work (Matthews et al 2011) has demonstrated that our single-cell RS techniques(Matthews et al 2010) applied to cells cultured and irradiated in vitro with single fractionsof ionizing radiation (15 30 and 50 Gy) can detect radiation-induced changes in proteinslipids and nucleic acids within a human prostate cancer cell line (DU145) In this workwe extend these methods to a panel of six human tumour cell lines derived from prostate(DU145 PC3 and LNCaP) breast (MDA-MB-231 and MCF7) and lung (H460) The celllines tested vary by tissue of origin p53 gene status and intrinsic radiosensitivity Within thetested cell lines we detect biochemical radiation response signatures that segregate accordingto radiosensitivity and p53 status For all cell lines our application of principal componentanalysis (PCA) effectively distinguishes radiation-induced biochemical changes from anybiochemical changes arising from cell cycle differences or other factors Our observedradiation response signatures indicate changes in the cellular concentration of aromatic aminoacids (tyrosine tryptophan and phenylalanine) conformational protein structures (α-helicesβ-sheets and random coils) and certain nucleic acid and lipid functional groups Theseradiation-induced biochemical changes are detected within the first 1ndash3 days after exposureto all doses delivered (15 30 and 50 Gy) Many of the observed biochemical changes areconsistent with known cellular response mechanisms to radiation exposure (eg synthesis anddegradation of structured proteins) whereas others may be novel discoveries The relationship


between the RS radiation response signatures and intrinsic radiosensitivity may be indicativeof the detection of biochemical mechanisms of radiation resistance or sensitivity with RSSome candidate processes for such mechanisms are discussed

2 Methods

21 Cell lines and culture conditions

Six human tumour cell lines were used for this study derived from prostate (DU145 PC3 andLNCaP) breast (MDA-MB-231 and MCF7) and lung (H460) These cell lines vary accordingto p53 status and average reported radiosensitivity (see section 35) given by the survivingfraction after 2 Gy (SF2) Our SF2 values were calculated from literature data from four orfive different laboratories per cell line (Eschrich et al 2009 Amorino et al 2000 Park et al2010 Sak et al 2002 Cai et al 2008 Robinson and Shewach 2001 Schmidt-Ullrich et al 1992Hahnel et al 2010 Wouters et al 2010 DeWeese et al 1998 Colletier et al 2000 Chendilet al 2004 Wang et al 2008 Fullerton et al 2004 Rosser et al 2004) whose methods closelyapproximated accepted protocols for clonogenic survival assays (Franken et al 2006)

Cells were cultured in a sterile environment using previously described protocols(Matthews et al 2010) and kept in an incubator at 5 CO2 and 37 C to promote growthCell stocks were sub-cultured every 3ndash4 days by rinsing the cells in phosphate buffered saline(PBS) (HyClone Logan UT) adding trypsin to detach the cells from the tissue culture flaskand transferring 10ndash20 of the harvested cells to a new flask containing fresh growth media

22 Cell irradiation

For all experiments ten identical cell cultures were prepared in T-75 flasks at an initialconfluency of 10ndash15 After sim35 days the culture media was replaced with fresh mediaand cultures were irradiated with a single fraction of 6 MV photons from a Varian 6EXlinear accelerator (Varian Medical Systems Inc Palo Alto CA USA) at a dose rate ofsim59 Gy minminus1 Four cultures were irradiated to 50 Gy one was irradiated to 15 Gy one wasirradiated to 30 Gy and the remaining four cultures were left unirradiated Cultures irradiatedto 50 Gy were harvested for RS analysis at 0 24 48 and 72 h post-irradiation with oneunirradiated culture harvested and analyzed at the same time Cultures irradiated to 15 and30 Gy were harvested and analyzed at 72 h post-irradiation

23 Cell cycle and viability analysis

During the harvesting procedure for RS analysis the cell cycle distribution and viability (livecell fraction) of each culture was determined with flow cytometry as described previously(Matthews et al 2010) For all experiments in this work the fraction of live cells in theharvested samples was between 75 and 99 Any cell spectrum resembling that of a deadcell (Notingher et al 2003) was rejected during RS acquisition No correlations were foundbetween the viability of harvested cultures and the RS data collected confirming that the cellspectra acquired for processing and PCA analysis were indeed obtained from live cells

24 RS and data processing

Sample preparation and RS analysis were performed as described previously (Matthews et al2010) Briefly after rinsing with PBS to remove dead cells and debris the remaining livecells were harvested with trypsin and centrifuged into a pellet in a 200 μL vial Vials


were kept on ice until RS analysis (1ndash6 h) upon which the chosen pellet was transferredto a quartz disk (Technical Glass Products Painesville OH USA) Raman spectra wereacquired from 20 individual cells from each sample with cells chosen at random from the toplayer of the cell pellet Raman acquisition was performed with an inVia Raman microscope(Renishaw Inc Hoffman Estates IL USA) with a 100times dry objective (Leica MicrosystemsWetzlar Germany) a 1200 lines mmminus1 diffraction grating 30 s acquisition time per cell and600ndash1800 cmminus1 spectral window A 785 nm continuous wave diode laser (Renishaw) was usedfor sample excitation providing a laser power density at the sample of sim05 mW μmminus3 Thesize of the sampling volume was sim2 times 5 times 10 μm allowing a single acquisition to representthe Raman spectrum of a single cell (sim10 μm diameter) (Matthews et al 2010) Each cellspectrum was processed to remove cosmic rays reduce noise via spectral smoothing estimateand subtract a baseline arising from the quartz substrate and biological fluorescence andnormalize to the total amount of biological material within the sampling volume (Matthewset al 2010) The fully processed data set (200 spectra per cell line) was then analyzed withPCA using standard algorithms (Matlab The Mathworks Natick MA USA) Correlations(r-values) between PCA components were computed using Pearsonrsquos linear correlationcoefficient (Matlab)

3 Results

31 Unirradiated cell spectra

The Raman spectrum of a single unirradiated DU145 cell from 600 to 1800 cmminus1 (figure 1(a))contains multiple contributions from proteins lipids and nucleic acids A detailed listing ofthe molecular assignments for the spectral features we observe for DU145 cells has beenrecently reported (Matthews et al 2010) compiled from literature reports examining differentcell lines or tissues (Notingher et al 2003 Notingher and Hench 2006 Krafft et al 2003Uzunbajakava et al 2003 Synytsya et al 2004 Omberg et al 2002 Borchman et al 1999)Here we also present the averaged Raman spectrum from 20 unirradiated cells for all six celllines used in this study (figure 1(b)) harvested and analyzed immediately after the time ofirradiation All spectral features observed for DU145 cells (figure 1(a)) are also observed forthe other cell lines There are subtle differences in the relative intensity of spectral featuresbetween cell lines arising from inherent differences in biomolecular compositions betweentumour cell lines (Crow et al 2005) and slightly different cell cycle distributions betweencultures at time of irradiation (Matthews et al 2010)

32 Cell cycle spectral variability

Our previous work using the prostate cell line DU145 demonstrated that Raman spectralvariability arising from inherent cell cycle differences between cells is identified by the firstPCA component of a RS data set obtained from both unirradiated and irradiated DU145 cells(Matthews et al 2011) In this study with multiple cell lines the cell cycle PCA componentobserved for DU145 cells (figure 1(c)) was consistently reproduced as either the first or secondPCA component of all data sets (figure 1(d)) and accounts for 10ndash50 (figure 1(d)) of the totalvariance in the data depending on the cell line This cell cycle variability arises from decreasedconcentrations of protein and nucleic acids (positive features in the PCA componentsfigures 1(c) and (d)) relative to lipids (negative features in the PCA componentsfigures 1(c) and (d)) in early G1 (or G0) phase cells as compared to late G1 S and G2 phasecells (Matthews et al 2010) It should be noted that the reduced percent variance explained


(a)

(b) (d)

(c)

Figure 1 (a) Sample Raman spectrum of a single unirradiated DU145 cell (b) Averagespectra from 20 unirradiated cells for the six cell lines used in this study (c) Cell cycle PCAcomponent for the DU145 data set (200 cells) (d) Cell cycle PCA components for all six cell lines(200 cells each) with percent variance explained by each component The Raman shift andmolecular origin of identifiable features are provided in (a) and (c) (Notingher et al 2003 Notingherand Hench 2006 Krafft et al 2003 Uzunbajakava et al 2003 Synytsya et al 2004 Omberg et al2002 Borchman et al 1999) Abbreviationsmdashp protein l lipid d DNARNA A adenine Tthymine G guanine C cytosine U uracil Phe phenylalanine Tyr tyrosine Trp tryptophanbk backbone def deformation tw twist sym symmetric asym asymmetric str stretch

by the H460 cell cycle component (121) relative to the other five cell lines (360ndash481)(figure 1(d)) is firstly a result of the H460 cell line exhibiting the largest radiation-inducedspectral changes (section 33) and secondly due to the fact that the H460 cell line demonstratedthe lowest levels of inherent susceptibility to depletion of nucleic acid and protein contentrelative to lipid content (via early G1 (or G0) arrest or like processes) induced by eitherradiation exposure or varying cell culture conditions (ie confluency)


Figure 2 Radiation-induced PCA components for all six cell lines with percent variance explainedby each component The Raman shift and molecular origin of identifiable features are provided infigures 3(a) (c) and (e)

33 RS radiation response signatures I categories R1 R2 and R3

PCA components corresponding to radiation-induced biochemical changes (independent fromthe cell cycle related changes described by the cell cycle PCA components figure 1(d)) wereidentified by statistically significant (p lt 005) changes in the corresponding PCA scoredistributions for irradiated samples as compared to unirradiated samples in the first 24ndash72h after irradiation In this study one definitively radiation-induced PCA component wasdetected for each cell line (figure 2) The percent variance explained by the radiation-inducedPCA component which indicates the strength of the radiation induced biochemical response(relative to other sources of spectral variability) and determines the PCA component numberis highly dependent upon cell line and varies from 30 (PC3) to 503 (H460) Correlationanalysis between radiation-induced PCA components (table 1) indicates that these RS radiationresponse signatures fall into three distinct categories which we abbreviate R1 (H460 andMCF7 r = 087) R2 (MDA-MB-231 and PC3 r = 059) and R3 (DU145 and LNCaPr = 051) There is also consistent correlation between the PCA components in groups R1


Table 1 Correlation r-values between radiation-induced PCA components (figure 2) for each cellline (1 = perfect correlation minus1 = perfect anti-correlation 0 = no correlation) All correlationvalues are statistically different than zero (p lt 005)

Cell line H460 MCF7 MDA-MB-231 PC3 DU145 LNCaP

H460 ndash 087 018 050 minus036 minus017MCF7 087 ndash 020 042 minus027 minus013MDA-MB-231 018 020 ndash 059 minus028 minus018PC3 050 042 059 ndash minus070 minus048DU145 minus036 minus027 minus028 minus070 ndash 051LNCaP minus017 minus013 minus018 minus048 051 ndash

and R2 (018 lt r lt 050) and consistent anti-correlation between the two PCA componentsin group R3 and the other four radiation-induced components (minus013 lt r lt minus070) Allcorrelations values between PCA components (table 1) are statistically different than zero(p lt 005)

34 RS radiation response signatures II radiation-induced changes in biomolecules acrosscategories R1 R2 and R3

The biomolecules responsible for the observed radiation-induced PCA components(figure 2) are identified by the positive and negative features in the PCA components Molecularassignments are provided (figure 3) for one cell line from each RS category using cell linesH460 (R1 figure 3(a)) MDA-MB-231 (R2 figure 3(c)) and DU145 (R3 figure 3(e)) asexamples The radiation-induced changes in these biomolecules are given by the changesin the corresponding PCA score distributions for irradiated cultures in the first 1ndash3 dayspost-irradiation relative to the unirradiated cultures (figures 3(b) (d) and (f)) For the PCAscore plots shown (figures 3(b) (d) and (f)) all irradiated samples demonstrated statisticallysignificant (p lt 005) decreases in their PCA scores from 24 to 72 h post-irradiation whencompared with unirradiated controls For the DU145 cell line (R3) both the radiation-inducedPCA component (figure 3(e)) and the corresponding PCA scores (figure 3(f)) match ourprevious observations (Matthews et al 2011)

For the PCA components from RS categories R1 and R2 (eg figures 3(a) and (c)) theobserved decrease in the PCA scores for irradiated cells (figure 3(b) and (d)) corresponds withincreases in amino acids α-helix protein structure and CH groups (common negative featuresin R1 and R2 PCA components) and with decreases in nucleic acids CH2 groups and β-sheetand random coil protein structures (common positive features in R1 and R2 PCA components)The presence of many common biomolecules in the PCA components between groups R1and R2 is reflected by the positive correlation between PCA components in these groups(table 1 average r = 033) For the RS category R3 components (eg figure 3(e)) the observeddecrease in the PCA scores for irradiated cells (figure 3(f)) corresponds with decreases in aminoacids α-helix protein structure and both CH2 and CH groups (common positive features inthe R3 PCA components) and with increases in nucleic acids β-sheet and random coilprotein structures and choline (common negative features in the R3 PCA components) Thepresence of many common biomolecules in the RS category R3 components that show oppositeradiation-induced changes compared to the R1 and R2 components is reflected by the negativecorrelation values between PCA components in group R3 and groups R1 and R2 (table 1minus013 lt r lt minus070)


(a) (b)

(c) (d)

(e) (f)

Figure 3 (a c e) Radiation-induced PCA components (a) H460 (c) MDA-MB-231 and(e) DU145 cell lines Raman shifts and molecular origins of identifiable peaks are provided(b d f) PCA scores for the (b) H460 (d) MDA-MB-231 and (f) DU145 radiation-induced PCAcomponents Different markers categorize all 200 cells by time of RS acquisition after irradiationThe average score and standard deviation is shown for each sample for visualization of the trendsin the data Abbreviationsmdashthe same as in figure 1

The PCA scores for the MCF7 (R1) PC3 (R2) and LNCaP (R3) radiation-induced PCAcomponents (supplementary figure S-1 available at stacksioporgPMB566839mmedia)show similar changes in their distributions with time and dose as their RS categorycounterparts For these cell lines there are occasional reductions in the distances between PCAscore distributions between irradiated and unirradiated samples (supplementary figures S-1b


Table 2 RS biochemical radiation response category tissue of origin (TOI) percent varianceexplained by radiation-induced PCA component G1 S and G2 fractions at 24 h post-irradiationp53 status and average reported radiosensitivity (SF2) for the six cell lines used in this study Thesuperscript numbers indicate literature references used which are specified below

RS Cell Variance G1 S G2 p53category line TOI () ( at 24 h) status SF2

R1 H460 Lung 503 73 10 17 wt[1] 064[2ndash5]

MCF7 Breast 73 40 9 51 wt[6] 064[237ndash9]

R2 MDA-MB-231 Breast 47 11 10 79 mt[6] 071[271011]

PC3 Prostate 30 7 6 87 mt[12] 064[213ndash15]

R3 DU145 Prostate 128 6 34 60 mt[12] 049[21316ndash18]

LNCaP Prostate 40 64 14 22 wt[12] 027[13141718]

[1] Mitchell et al (2010) [2] Eschrich et al (2009) [3] Amorino et al (2000) [4] Park et al (2010)[5] Sak et al (2002) [6] Hui et al (2006) [7] Cai et al (2008) [8] Robinson and Shewach (2001)[9] Schmidt-Ullrich et al (1992) [10] Hahnel et al (2010) [11] Wouters et al (2010) [12] Williams et al(2008) [13] DeWeese et al (1998) [14] Colletier et al (2000) [15] Chendil et al (2004) [16] Wanget al (2008) [17] Fullerton et al (2004) [18] Rosser et al (2004)

S-1d and S-1f available at stacksioporgPMB566839mmedia) a result consistent with thereduced strength of the radiation-induced responses of these cell lines relative to their RScategory counterparts (figure 2)

35 RS radiation response associations with cell cycle arrest p53 and radiosensitivity

The RS radiation response categories R1 R2 and R3 segregate according to the knownradiosensitivity of the cell lines (table 2) The R1 and R2 category cell lines are known tobe radiation resistant (SF2 gt 06) and the R3 cell lines are comparatively radiation sensitive(SF2 lt 05) Furthermore the R1 cell lines contain a wild-type (wt) p53 gene whereas theR2 cells contain a mutant (mt) p53 gene The two R3 cell lines DU145 (mt p53) and LNCaP(wt p53) are radiosensitive and show very different biochemical radiation response signaturesfrom their p53 counterparts (figure 2 table 1)

As normally functioning wt p53 is required for G1 phase cell cycle arrest post-irradiation(McIlwrath et al 1994) our measured fraction of G1 cells at 24 h post-irradiation (table 2)confirms that the mt p53 cell lines do not show G1 phase arrest (6ndash11 G1 fraction at 24 hpost-irradiation) and in fact show high levels of radiation-induced G2 phase arrest (60ndash87G2 fraction at 24 h post-irradiation) As expected the wt p53 cell lines maintain high G1fractions post-irradiation (40ndash73 G1 fraction at 24 h post-irradiation)

4 Discussion

41 RS detection of biochemical signatures of radiation response

Our methods used in this study namely the acquisition of hundreds of high-quality single-cell RS spectra per cell line established spectral processing techniques (Matthews et al2010 2011) and PCA enable us to separate radiation-induced spectral changes from othersimultaneously occurring sources of spectral variability such as cell cycle This techniqueprovides direct analysis of the biomolecular changes arising in single cells responding to


radiation exposure independent of cell cycle or cell-death-related processes We identifyRS signatures of radiation response (ie the radiation-induced PCA components figure 2)by statistically significant shifts in the corresponding PCA score distributions (eg figures3(b) (d) and (f)) as a function of time post-irradiation and (in some cell lines) the delivereddose The biomolecules responsible for these signatures are identified by the known molecularassignments of the positive and negative features in the radiation-induced PCA components(figures 1(a) 3(a) (c) and (e)) In this study we find that the biomolecular radiation responsesof the six cell lines segregate into distinct categories (R1 R2 and R3) observable bothby visual inspection of the signatures (figure 2) and by correlation analysis (table 1) Thequalitative similarities observed in the biomolecular changes between categories R1 and R2and the many opposite changes observed between category R3 and categories R1 and R2(ie figures 3(a) (c) and (e)) are quantitatively confirmed via correlation analysis (table 1)

42 Segregation of common radiation response signatures according to p53 status andradiosensitivity

The four cell lines that fall into the RS radiation response categories R1 and R2 are known tobe radiation resistant (SF2 gt 06) whereas the two cell lines in category R3 are comparativelyradiation sensitive (SF2 lt 05) (table 2) The RS signatures of radiation response betweencategories R1 (resistant wt p53) and R2 (resistant mt p53) are different but share manycommon molecular features (figure 2) indicating similar yet unique radiation responses Asboth groups R1 and R2 are comprised solely of radiation resistant cell lines it is possiblethat the biochemical radiation responses we observe with RS are caused by cellular responsemechanisms that increase survival after radiation exposure Candidate mechanisms for suchresponses are discussed below (section 43)

The available SF2 data from different laboratories (eg for MCF7 SF2 values rangefrom 050 (Schmidt-Ullrich et al 1992) to 081 (Amorino et al 2000)) separate the confirmedresistant cell lines (average SF2 gt 06) from the comparatively sensitive cell lines (averageSF2 lt 05) The average SF2 values used here (table 2) were calculated from 4ndash5 literaturesources that closely approximated currently accepted protocols for clonogenic survival assays(Franken et al 2006) In particular it was ensured that cultures were irradiated in exponentialgrowth phase and that experiments were performed at least in triplicate

It is important to note that PCA also calculates the relative lsquostrengthrsquo of the RS observedbiomolecular radiation response given by the variance explained by the identified radiation-induced PCA component In our study the cell lines demonstrated different radiation responsestrengths between categories of radiation response and also within each category (figure 2table 2) Between categories R1 and R2 the resistant wt p53 cell lines (R1) respondedstronger than the resistant mt p53 cell lines (R2) A possible explanation for this behavior isproposed below (section 431) Within each category H460 responded stronger than MCF7(R1) MDA-MB-231 responded stronger than PC3 (R2) and DU145 responded stronger thanLNCaP (R3)

43 Biochemical mechanisms of radiation resistance or sensitivity

431 Radioresistant cell lines RS categories R1 and R2 Some of the most dramaticradiation-induced biochemical changes we observe with RS arise from proteins (figure 3)From 24 to 72 h after irradiation the known radiation resistant cell lines (RS categoriesR1 and R2) demonstrate increased concentrations of aromatic amino acids (phenylalaninetyrosine and tryptophan) and α-helix protein structures and decreased concentrations ofβ-sheet and random coil protein structures relative to unirradiated controls These changes


corroborate with prior evidence that synthesis and degradation of structured proteins iscorrelated with increased survival post-irradiation and plays an important role in cellularradioadaptive response (Tapio and Jacob 2007) These changes may also be reflective ofcellular survival mechanisms triggered by radiation-induced oxidative stress involving thebreakdown of structured proteins (ie β-sheet and random coil protein structures) into freeamino acids to aid in the scavenging of reactive oxygen species created by radiation damage(Droge 2002) Our observation of increased α-helix protein structure with radiation could beexplained by concurrent synthesis of certain proteins involved in radiation response pathwaysthat increase cell survival post-irradiation For example colorectal cells have been shown toexpress 14-3-3σ protein for up to 60 h post-irradiation to aid in the inhibition of cell cycleprogression through G2M phase resulting in increased survival (Hermeking et al 1997)14-3-3 proteins are primarily composed of α-helices (Xiao et al 1995) are involved in manycell division and signaling pathways (Xiao et al 1995 van Hemert et al 2001) and are knownto suppress apoptosis via inhibition of several pro-apoptosis pathways (van Hemert et al2001) Another candidate protein with predominantly α-helix structure is survivin (Chantalatet al 2000) a known anti-apoptosis factor that has been linked to increased radioresistance inglioblastomas (Chakravarti et al 2004) pancreatic cancers (Kami et al 2005) rectal cancers(Rodel et al 2005) and head and neck squamous carcinomas (Khan et al 2010) Furthermoreit was shown for glioblastomas that radioresistant cell lines expressed survivin post-irradiationin all phases of the cell cycle whereas radiosensitive cell lines limited survivin expression toG2M phase independent of p53 status (Chakravarti et al 2004)

In this study we observe unique radiation response signatures between the resistant wtp53 (R1) and the resistant mt p53 (R2) cell lines However the consistent positive correlationsbetween the R1 and R2 radiation response signatures (table 1 018 lt r lt 050) indicatethat there are similarities in the radiation-induced biomolecular responses between resistantwt p53 and resistant mt p53 cell lines p53 is known to regulate cell cycle arrest post-irradiation (McIlwrath et al 1994 Hermeking et al 1997) and our flow cytometry measurements(table 2) confirm the differences in cell cycle regulation that occur between the wt p53 andthe mt p53 cell lines in our study Furthermore it has been observed that radiation-inducedapoptosis which would increase sensitivity requires wt p53 status (McIlwrath et al 1994)As such it is to be expected that the biochemical nature of a radiation-induced response thatpromotes cell survival post-irradiation may be determined by p53 status An example of suchp53 dependence on biochemical radiation response and radiosensitivity has been recentlyobserved via analysis of micro-RNA expression post-irradiation (Chaudhry et al 2010) It isalso likely that different responses working to achieve the same result (ie increased survivalpost-irradiation) would have common characteristics (eg expression of anti-apoptosis factorsor other survival signals) As discussed above in section 42 we observe stronger radiationresponses from the R1 (resistant wt p53) cell lines than from the R2 (resistant mt p53) celllines although the responses have many similar features If our observed RS responses fromthe R1 and R2 cell lines are indeed caused in part by the radiation-induced synthesis of anti-apoptosis proteins it is plausible that a wt p53 cell line with equivalent radiosensitivity as amt p53 cell line would need to mount a larger biochemical response (ie increased expressionof anti-apoptosis factors) in order to suppress the intact pro-apoptotic pathways in additionto the biochemical response required for other mechanisms of survival post-irradiation Thismodel may explain the observed differences and similarities between the radiation responsesignatures obtained for groups R1 and R2 in this study

432 Radiosensitive cell lines RS category R3 Neither of the radiosensitive cell linescomprising RS category R3 DU145 (mt p53) and LNCaP (wt p53) show a radiation response


similar to the other four cell lines This lack of response may be the result of some other reasonwhy these cells do not mount a similar radiation response as their p53 counterparts which maybe necessary for increased survival post-irradiation (eg expression of anti-apoptosis factorsor other survival signals) Furthermore both of the radiation response signatures for the R3cell lines although different (figure 2) are anti-correlated with the R1 and R2 signatures andpositively correlated with each other (table 1) As such these R3 radiation response signatureswhich show anti-correlation with possible biochemical signatures of radiation resistance mayin turn prove to be signatures of radiation sensitivity

44 Uniqueness of the observed RS biochemical radiation responses

The biochemical variability described by the radiation-induced PCA components (figure 2)is only observed when the data sets containing both unirradiated and irradiated cell spectraare input into PCA with one exception The radiation-induced PCA component for theH460 cell line which demonstrated the strongest radiation response of all the cell lines(503 of the total variance) is also observed (r = 080) when only the unirradiatedH460 cell spectra are input into PCA but with a much reduced percent variance explained(168 versus 503) This result suggests that the biochemical variability described bythe radiation-induced PCA component for the H460 cell line is not a uniquely radiation-induced response but rather is enhanced by radiation and may be induced by other factorsas well Interestingly a large fraction of the variability described by the component obtainedfrom only the unirradiated cells data set arises from a small number of outliers (sim4) fromthe 80 unirradiated cells The corresponding PCA component and score plot obtained fromonly the 80 unirradiated H460 cells are provided in supplementary figure S-2 available atstacksioporgPMB566839mmedia where the four outlier cells with the lowest PCA scoresare indicated If these four outliers are removed from the 80 cell data set prior to PCA thecomponent is still observed but with a significant drop in variance (111 versus 168)

For the other five human tumour cell lines investigated in this work the radiation-inducedPCA components are not observed if only the unirradiated cell spectra are input into PCAIf the biochemical variability was indeed present in these other cell lines it was too weakfor detection with RS using the current methods possibly due to insufficient induction by thein vitro culture environment Since the radiation response of the H460 cell line is inherentlymuch stronger than that of the other cell lines it is possible that the in vitro culture environmentprovided sufficient stimuli or stresses to induce a similar biochemical response in a subset ofthe unirradiated H460 cells which was therefore detected with RS without radiation exposure

45 The effect of radiation on cell cycle variability

We previously reported for RS of irradiated DU145 cells how radiation induces an observableeffect on the PCA scores for the cell cycle PCA component while leaving the features ofthe cell cycle PCA component unchanged (Matthews et al 2011) Here we report the sameresult for all six cell lines used in this study (figure 1(d) and table 2) The effect of radiationon the corresponding PCA scores (not shown) is dependent on the susceptibility of the cellline to radiation-induced depletion of nucleic acid and protein content relative to lipid contentarising from early G1 (or G0) arrest or like processes (Matthews et al 2010) Determining anypossible relationships between this susceptibility and the nature of the observed RS signaturesof radiation response or with known radiosensitivity may be a topic of interest for futureresearch and will require similar analysis of more cell lines


5 Conclusions

Within a preliminary panel of six human tumour cell lines derived from prostate (DU145 PC3and LNCaP) breast (MDA-MB-231 and MCF7) and lung (H460) we have demonstratedthat RS can detect biochemical signatures of in vitro radiation response that segregateaccording to p53 status and intrinsic radiosensitivity (SF2) The observed RS signaturesarise from radiation-induced changes in cellular concentrations of aromatic amino acidsconformational protein structures and certain nucleic acid and lipid functional groups and aredetected from live unfixed single cells analyzed 1ndash3 days post-irradiation Our sensitivityto the biomolecules responsible for the observed radiation responses provides new insightinto possible mechanisms of radiation survival and into the differences in such survivalmechanisms between wt p53 and mt p53 cell lines We have proposed potential radiation-induced biochemical response mechanisms underlying our RS observations namely (1) theregulated synthesis and degradation of structured proteins and (2) the expression of anti-apoptosis factors or other survival signals

This study further demonstrates the utility of using RS for radiobiological investigationsSpecifically the relationship between the RS radiation response signatures and intrinsicradiosensitivity supports the possibility of using RS for detecting radiation resistance orsensitivity in clinical practice Future work may lead to the development of RS techniques formonitoring or predicting tumour response in radiation therapy patients

Acknowledgments

The authors gratefully acknowledge funding from the National Science and EngineeringResearch Council the Canadian Foundation for Innovation and the Western EconomicDiversification program We would also like to thank the staff of the Deeley ResearchCentre at the BC Cancer Agencyrsquos Vancouver Island Centre for providing cell stocks facilitiesand workspace

References

Amorino G Freeman M and Choy H 2000 Enhancement of radiation effects in vitro by the estrogen metabolite2-methoxyestradiol Radiat Res 153 384ndash91

Begg A et al 1999 The value of pretreatment cell kinetic parameters as predictors for radiotherapy outcome in headand neck cancer a multicenter analysis Radiother Oncol 50 13ndash23

Bjork-Eriksson T West C Karlsson E and Mercke C 2000 Tumor radiosensitivity (SF2) is a prognostic factor forlocal control in head and neck cancers Int J Radiat Oncol Biol Phys 46 13ndash9

Borchman D Tang D and Yappert M 1999 Lipid composition membrane structure relationships in lens and musclesarcoplasmic reticulum membranes Biospectroscopy 5 151ndash67

Cai Z Chen Z Bailey K Scollard D Reilly R and Vallis K 2008 Relationship between induction of phosphorylatedH2AX and survival in breast cancer cells exposed to 111In-DTPA-hEGF J Nucl Med 49 1353ndash61

Chakravarti A Zhai G Zhang M Malhotra R Latham D Delaney M Robe P Nestler U Song Q and LoefflerJ 2004 Survivin enhances radiation resistance in primary human glioblastoma cells via caspase-independentmechanisms Oncogene 23 7494ndash506

Chantalat L Skoufias D Kleman J Jung B Dideberg O and Margolis R 2000 Crystal structure of human survivinreveals a bow tie-shaped dimer with two unusual alpha-helical extensions Mol Cell 6 183ndash9

Chaudhry M Kreger B and Omaruddin R 2010 Transcriptional modulation of micro-RNA in human cells differingin radiation sensitivity Int J Radiat Biol 86 569ndash83

Chendil D Ranga R Meigooni D Sathishkumar S and Ahmed M 2004 Curcumin confers radiosensitizing effect inprostate cancer cell line PC-3 Oncogene 23 1599ndash607


Choi J Choo J Chung H Gweon D Park J Kim H Park S and Oh C 2005 Direct observation of spectraldifferences between normal and basal cell carcinoma (BCC) tissues using confocal Raman microscopyBiopolymers 77 264ndash72

Colletier P Ashoori F Cowen D Meyn R Tofilon P Meistrich M and Pollack A 2000 Adenoviral-mediated p53transgene expression sensitizes both wild-type and null p53 prostate cancer cells in vitro to radiation Int JRadiat Oncol Biol Phys 48 1507ndash12

Crow P Barrass B Kendall C Hart-Prieto M Wright M Persad R and Stone N 2005 The use of Raman spectroscopyto differentiate between different prostatic adenocarcinoma cell lines Brit J Cancer 92 2166ndash70

de Jong B Bakker T Maquelin K van der Kwast T Bangma C Kok D and Puppels G 2006 Discrimination betweennontumor bladder tissue and tumor by Raman spectroscopy Anal Chem 78 7761ndash9

DeWeese T Shipman J Dillehay L and Nelson W 1998 Sensitivity of human prostatic carcinoma cell lines to lowdose rate radiation exposure J Urol 159 591ndash8

Droge W 2002 Free radicals in the physiological control of cell function Physiol Rev 82 47ndash95Eschrich S et al 2009 A gene expression model of intrinsic tumor radiosensitivity prediction of response and

prognosis after chemoradiation Int J Radiat Oncol Biol Phys 75 489ndash96Eschrich S Zhang H Zhao H Boulware D Lee J Bloom G and Torres-Roca J 2009 Systems biology modeling

of the radiation sensitivity network a biomarker discovery platform Int J Radiat Oncol Biol Phys75 497ndash505

Franken N Rodermond H Stap J Haveman J and van Bree C 2006 Clonogenic assay of cells in vitro NatProtoc 1 2315ndash9

Fullerton N Boyd M Mairs R Keith W Alderwish O Brown M Livingstone A and Kirk D 2004 Combining atargeted radiotherapy and gene therapy approach for adenocarcinoma of prostate Prostate Cancer ProstaticDis 7 355ndash63

Hahnel A Wichmann H Kappler M Kotzsch M Vordermark D Taubert H and Bache M 2010 Effects of osteopontininhibition on radiosensitivity of MDA-MB-231 breast cancer cells Radiat Oncol 5 82

Hermeking H Lengauer C Polyak K He T Zhang L Thiagalingam S Kinzler K and Vogelstein B 1997 14-3-3sigma is a p53-regulated inhibitor of G2M progression Mol Cell 1 3ndash11

Hui L Zheng Y Yan Y Bargonetti J and Foster D 2006 Mutant p53 in MDA-MB-231 breast cancer cells isstabilized by elevated phospholipase D activity and contributes to survival signals generated by phospholipaseD Oncogene 25 7305ndash10

Kami K et al 2005 Downregulation of survivin by siRNA diminishes radioresistance of pancreatic cancer cellsSurgery 138 299ndash305

Khan Z Khan N Tiwari R Patro I Prasad G and Bisen P 2010 Down-regulation of survivin by oxaliplatin diminishesradioresistance of head and neck squamous carcinoma cells Radiother Oncol 96 267ndash73

Krafft C Knetschke T Siegner A Funk R and Salzer R 2003 Mapping of single cells by near infrared Ramanmicrospectroscopy Vib Spectrosc 32 75ndash83

Kunapareddy N Freyer J and Mourant J 2008 Raman spectroscopic characterization of necrotic cell death J BiomedOpt 13 054002

Lakshmi R Kartha V Krishna C Solomon J Ullas G and Devi P 2002 Tissue Raman spectroscopy for the study ofradiation damage brain irradiation of mice Radiat Res 157 175ndash82

Levine E et al 1995 Apoptosis intrinsic radiosensitivity and prediction of radiotherapy response in cervical-carcinomaRadiother Oncol 37 1ndash9

Lieber C and Mahadevan-Jansen A 2003 Automated method for subtraction of fluorescence from biological Ramanspectra Appl Spectrosc 57 1363ndash7

Lieber C Majumder S Billheimer D Ellis D and Mahadevan Jansen A 2008 Raman microspectroscopy for skincancer detection in vitro J Biomed Opt 13 024013

Luukkaa M Jokilehto T Kronqvist P Vahlberg T Grenman R Jaakkola P and Minn H 2009 Expression of thecellular oxygen sensor PHD2 (EGLN-1) predicts radiation sensitivity in squamous cell cancer of the head andneck Int J Radiat Biol 85 900ndash8

Matthews Q Brolo A Lum J Duan X and Jirasek A 2011 Raman spectroscopy of single human tumour cells exposedto ionizing radiation in vitro Phys Med Biol 56 19ndash38

Matthews Q Jirasek A Lum J Duan X and Brolo A 2010 Variability in Raman spectra of single human tumor cellscultured in vitro correlation with cell cycle and culture confluency Appl Spectrosc 64 871ndash87

McIlwrath A Vasey P Ross G and Brown R 1994 Cell-cycle arrests and radiosensitivity of human tumor-cell linesdependence on wild-type p53 for radiosensitivity Cancer Res 54 3718ndash22

Mitchell J Choudhuri R Fabre K Sowers A Citrin D Zabludoff S and Cook J 2010 In vitro and in vivoradiation sensitization of human tumor cells by a novel checkpoint kinase inhibitor AZD7762 Clin CancerRes 16 2076ndash84


Motz J Hunter M Galindo L Gardecki J Kramer J Dasari R and Feld M 2004 Optical fiber probe for biomedicalRaman spectroscopy Appl Opt 43 542ndash54

Mourant J Dominguez J Carpenter S Short K Powers T Michalczyk R Kunapareddy N Guerra A and Freyer J2006 Comparison of vibrational spectroscopy to biochemical and flow cytometry methods for analysis of thebasic biochemical composition of mammalian cells J Biomed Opt 11 064024

Nijssen A Schut T Heule F Caspers P Hayes D Neumann M and Puppels G 2002 Discriminating basal cellcarcinoma from its surrounding tissue by Raman spectroscopy J Invest Dermatol 119 64ndash9

Nordsmark M and Overgaard J 2000 A confirmatory prognostic study on oxygenation status and loco-regional controlin advanced head and neck squamous cell carcinoma treated by radiation therapy Radiother Oncol 57 39ndash43

Notingher I and Hench L 2006 Raman microspectroscopy a noninvasive tool for studies of individual living cellsin vitro Expert Rev Med Devices 3 215ndash34

Notingher I Verrier S Haque S Polak J and Hench L 2003 Spectroscopic study of human lung epithelial cells (A549)in culture living cells versus dead cells Biopolymers 72 230ndash40

Notingher I Verrier S Romanska H Bishop A Polak J and Hench L 2002 In situ characterisation of living cells byRaman spectroscopy Spectroscopy 16 43ndash51

Omberg K Osborn J Zhang S Freyer J Mourant J and Schoonover J 2002 Raman spectroscopy and factor analysisof tumorigenic and non-tumorigenic cells Appl Spectrosc 56 813ndash9

Oya N Zolzer F Werner F and Streffer C 2003 Effects of serum starvation on radiosensitivity proliferation andapoptosis in four human tumor cell lines with different p53 status Strahlenther Onkol 179 99ndash106

Park S Kim Y and Pyo H 2010 Gefitinib radiosensitizes non-small cell lung cancer cells through inhibition of ataxiatelangiectasia mutated Mol Cancer 9 222

Peters L 1996 Radiation therapy tolerance limitsmdashfor one or for all Janeway lecture Cancer 77 2379ndash85Puppels G Olminkhof J Segersnolten G Otto C Demul F and Greve J 1991 Laser irradiation and Raman spectroscopy

of single living cells and chromosomes sample degradation occurs with 5145 nm but not with 660 nm laserlight Exp Cell Res 195 361ndash7

Robinson B and Shewach D 2001 Radiosensitization by gemcitabine in p53 wild-type and mutant MCF-7 breastcarcinoma cell lines Clin Cancer Res 7 2581ndash9

Rodel F Hoffmann J Distel L Herrmann M Noisternig T Papadopoulos T Sauer R and Rodel C 2005 Survivinas a radioresistance factor and prognostic and therapeutic target for radiotherapy in rectal cancer CancerRes 65 4881ndash7

Rosser C Tanaka M Pisters L Tanaka N Levy L Hoover D Grossman H Mcdonnell T Kuban D and Meyn R 2004Adenoviral-mediated PTEN transgene expression sensitizes Bcl-2-expressing prostate cancer cells to radiationCancer Gene Ther 11 273ndash9

Sailer K Viaggi S and Nusse M 1996 Radiation-induced structural modifications in dsDNA analysed by FT-Ramanspectroscopy Int J Radiat Biol 69 601ndash13

Sailer K Viaggi S and Nusse M 1997 Kinetics of radiation- and cytochrome c-induced modifications in liposomesanalysed by FT-Raman spectroscopy Biochim Biophys Acta 1329 259ndash68

Sak A Stuschke M Wurm R Schroeder G Sinn B Wolf G and Budach V 2002 Selective inactivation of DNA-dependent protein kinase with antisense oligodeoxynucleotides consequences for the rejoining of radiation-induced DNA double-strand breaks and radiosensitivity of human cancer cell lines Cancer Res 62 6621ndash4

Schmidt-Ullrich R Valerie K Chan W Wazer D and Lin P 1992 Expression of oestrogen receptor and transforminggrowth factor-alpha in MCF-7 cells after exposure to fractionated irradiation Int J Radiat Biol 61 405ndash15

Schulze G Jirasek A Yu M Lim A Turner R and Blades M 2005 Investigation of selected baseline removal techniquesas candidates for automated implementation Appl Spectrosc 59 545ndash74

Scott S Earle J and Gumerlock P 2003 Functional p53 increases prostate cancer cell survival after exposure tofractionated doses of ionizing radiation Cancer Res 63 7190ndash6

Shaw C and Jirasek A 2009 The use of ultraviolet resonance Raman spectroscopy in the analysis of ionizing-radiation-induced damage in DNA Appl Spectrosc 63 412ndash22

Short K Carpenter S Freyer J and Mourant J 2005 Raman spectroscopy detects biochemical changes due toproliferation in mammalian cell cultures Biophys J 88 4274ndash88

Swain R Jell G and Stevens M 2008 Non-invasive analysis of cell cycle dynamics in single living cells with Ramanmicro-spectroscopy J Cell Biochem 104 1427ndash38

Synytsya A et al 2004 Raman spectroscopy of tissue samples irradiated by protons Int J Radiat Biol 80 581ndash91Synytsya A Synytsya A Alexa P Wagner R Davıdkova M and Volka K 2011 Raman spectroscopic study on sodium

hyaluronate an effect of proton and γ irradiation J Raman Spectrosc 42 544ndash50Tapio S and Jacob V 2007 Radioadaptive response revisited Radiat Environ Biophys 46 1ndash12Teh S Zheng W Ho K Teh M and Yeoh K 2009 Near-infrared Raman spectroscopy for gastric precancer diagnosis

J Raman Spectrosc 40 908ndash14


Teh S Zheng W Ho K Teh M Yeoh K and Huang Z 2008 Diagnosis of gastric cancer using near-infrared Ramanspectroscopy and classification and regression tree techniques J Biomed Opt 13 034013

Tollefson M Magera J Sebo T Cohen J Drauch A Maier J and Frank I 2010 Raman spectral imaging of prostatecancer can Raman molecular imaging be used to augment standard histopathology BJU Int 106 484ndash8

Torres-Roca J et al 2005 Prediction of radiation sensitivity using a gene expression classifier Cancer Res 65 7169ndash76Uzunbajakava N Lenferink A Kraan Y Willekens B Vrensen G Greve J and Otto C 2003 Nonresonant Raman

imaging of protein distribution in single human cells Biopolymers 72 1ndash9van Hemert M Steensma H and van Heusden G 2001 14-3-3 proteins key regulators of cell division signalling and

apoptosis BioEssays 23 936ndash46Vaupel P and Mayer A 2007 Hypoxia in cancer significance and impact on clinical outcome Cancer Metastasis

Rev 26 225ndash39Verma S 1986 Low-levels of irradiation modify lipid domains in model membranesmdasha laser Raman study Radiat

Res 107 183ndash93Verma S and Rastogi A 1990 Role of proteins in protection against radiation-induced damage in membranes Radiat

Res 122 130ndash6Verma S Singhal A and Sonwalkar N 1993 Ionizing-radiation target groups of band-3 inserted into egg lecithin

liposomes as determined by Raman spectroscopy Int J Radiat Biol 63 279ndash88Verma S and Sonwalkar N 1991 Structural changes in plasma membranes prepared from irradiated Chinese-hamster

V79-cells as revealed by Raman-spectroscopy Radiat Res 126 27ndash35Verrier S Notingher I Polak J and Hench L 2004 In situ monitoring of cell death using Raman microspectroscopy

Biopolymers 74 157ndash62Vidyasagar M Maheedhar K Vadhiraja B Fernendes D Kartha V and Krishna C 2008 Prediction of radiotherapy

response in cervix cancer by Raman spectroscopy a pilot study Biopolymers 89 530ndash7Wang J Rhee J Shi P Stewart R and Li X A 2008 In vitro determination of radiation sensitivity parameters for

DU-145 prostate cancer cells Int J Radiat Biol 84 515ndash22West C Davidson S Roberts S and Hunter R 1997 The independence of intrinsic radiosensitivity as a prognostic

factor for patient response to radiotherapy of carcinoma of the cervix Brit J Cancer 76 1184ndash90Williams J Zhang Y Russell J Koch C and Little J 2007 Human tumor cells segregate into radiosensitivity groups

that associate with ATM and TP53 status Acta Oncol 46 628ndash38Williams J Zhang Y Zhou H Gridley D Koch C Russell J Slater J and Little J 2008 A quantitative overview of

radiosensitivity of human tumor cells across histological type and TP53 status Int J Radiat Biol 84 253ndash64Williams J Zhang Y Zhou H Gridley D Koch C Slater J and Little J 2008 Overview of radiosensitivity of human

tumor cells to low-dose-rate irradiation Int J Radiat Oncol Biol Phys 72 909ndash17Wouters A Pauwels B Lambrechts H Pattyn G Ides J Baay M Meijnders P Lardon F and Vermorken J

2010 Counting clonogenic assays from normoxic and anoxic irradiation experiments manually or by usingdensitometric software Phys Med Biol 55 N167ndash78

Xiao B Smerdon S Jones D Dodson G Soneji Y Aitken A and Gamblin S 1995 Structure of a 14-3-3 protein andimplications for coordination of multiple signalling pathways Nature 376 188ndash91

Zoladek A Pascut F Patel P and Notingher I 2011 Non-invasive time-course imaging of apoptotic cells by confocalRaman micro-spectroscopy J Raman Spectrosc 42 251ndash8

1 Introduction

2 Methods


22 Cell irradiation



3 Results




34 RS radiation response signatures II radiation-induced changes in biomolecules across categories R1 R2 and R3


4 Discussion


42 Segregation of common radiation response signatures according to p53 status and radiosensitivity




5 Conclusions

IOP PUBLISHING PHYSICS IN MEDICINE AND BIOLOGY

Phys Med Biol 56 (2011) 6839ndash6855 doi1010880031-91555621006

Biochemical signatures of in vitro radiation responsein human lung breast and prostate tumour cellsobserved with Raman spectroscopy

Q Matthews1 A Jirasek1 J J Lum2 and A G Brolo3

1 Department of Physics and Astronomy University of Victoria Victoria BC V8W 3P6 Canada2 Deeley Research Centre BC Cancer Agency Vancouver Island Centre Victoria BC V8R 6V5Canada3 Department of Chemistry University of Victoria Victoria BC V8W 3V6 Canada

E-mail qmatthewuvicca and jirasekuvicca

Received 7 July 2011 in final form 5 September 2011Published 5 October 2011Online at stacksioporgPMB566839

AbstractThis work applies noninvasive single-cell Raman spectroscopy (RS) andprincipal component analysis (PCA) to analyze and correlate radiation-inducedbiochemical changes in a panel of human tumour cell lines that vary by tissueof origin p53 status and intrinsic radiosensitivity Six human tumour cell linesderived from prostate (DU145 PC3 and LNCaP) breast (MDA-MB-231 andMCF7) and lung (H460) were irradiated in vitro with single fractions (15 30 or50 Gy) of 6 MV photons Remaining live cells were harvested for RS analysisat 0 24 48 and 72 h post-irradiation along with unirradiated controls Single-cell Raman spectra were acquired from 20 cells per sample utilizing a 785 nmexcitation laser All spectra (200 per cell line) were individually post-processedusing established methods and the total data set for each cell line was analyzedwith PCA using standard algorithms One radiation-induced PCA componentwas detected for each cell line by identification of statistically significantchanges in the PCA score distributions for irradiated samples as compared tounirradiated samples in the first 24ndash72 h post-irradiation These RS responsesignatures arise from radiation-induced changes in cellular concentrations ofaromatic amino acids conformational protein structures and certain nucleicacid and lipid functional groups Correlation analysis between the radiation-induced PCA components separates the cell lines into three distinct RS responsecategories R1 (H460 and MCF7) R2 (MDA-MB-231 and PC3) and R3(DU145 and LNCaP) These RS categories partially segregate according toradiosensitivity as the R1 and R2 cell lines are radioresistant (SF2 gt 06) andthe R3 cell lines are radiosensitive (SF2 lt 05) The R1 and R2 cell lines furthersegregate according to p53 gene status corroborated by cell cycle analysis post-irradiation Potential radiation-induced biochemical response mechanismsunderlying our RS observations are proposed such as (1) the regulated synthesis

0031-915511216839+17$3300 copy 2011 Institute of Physics and Engineering in Medicine Printed in the UK 6839




1 Introduction









2 Methods




22 Cell irradiation








3 Results






(a)

(b) (d)

(c)
















(a) (b)

(c) (d)

(e) (f)






R1 H460 Lung 503 73 10 17 wt[1] 064[2ndash5]


R2 MDA-MB-231 Breast 47 11 10 79 mt[6] 071[271011]









4 Discussion























5 Conclusions



Acknowledgments


References






















































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions




1 Introduction









2 Methods




22 Cell irradiation








3 Results






(a)

(b) (d)

(c)
















(a) (b)

(c) (d)

(e) (f)






R1 H460 Lung 503 73 10 17 wt[1] 064[2ndash5]


R2 MDA-MB-231 Breast 47 11 10 79 mt[6] 071[271011]









4 Discussion























5 Conclusions



Acknowledgments


References






















































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions







2 Methods




22 Cell irradiation








3 Results






(a)

(b) (d)

(c)
















(a) (b)

(c) (d)

(e) (f)






R1 H460 Lung 503 73 10 17 wt[1] 064[2ndash5]


R2 MDA-MB-231 Breast 47 11 10 79 mt[6] 071[271011]









4 Discussion























5 Conclusions



Acknowledgments


References






















































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions



3 Results






(a)

(b) (d)

(c)
















(a) (b)

(c) (d)

(e) (f)






R1 H460 Lung 503 73 10 17 wt[1] 064[2ndash5]


R2 MDA-MB-231 Breast 47 11 10 79 mt[6] 071[271011]









4 Discussion























5 Conclusions



Acknowledgments


References






















































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions


(a)

(b) (d)

(c)
















(a) (b)

(c) (d)

(e) (f)






R1 H460 Lung 503 73 10 17 wt[1] 064[2ndash5]


R2 MDA-MB-231 Breast 47 11 10 79 mt[6] 071[271011]









4 Discussion























5 Conclusions



Acknowledgments


References






















































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions














(a) (b)

(c) (d)

(e) (f)






R1 H460 Lung 503 73 10 17 wt[1] 064[2ndash5]


R2 MDA-MB-231 Breast 47 11 10 79 mt[6] 071[271011]









4 Discussion























5 Conclusions



Acknowledgments


References






















































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions




R1 H460 Lung 503 73 10 17 wt[1] 064[2ndash5]


R2 MDA-MB-231 Breast 47 11 10 79 mt[6] 071[271011]









4 Discussion























5 Conclusions



Acknowledgments


References






















































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions





















5 Conclusions



Acknowledgments


References






















































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions













































































1 Introduction

2 Methods


22 Cell irradiation



3 Results






4 Discussion






5 Conclusions

biochemical signatures of radiation response in tumour ...agbrolo/pmb_quinn_2011_nov.pdf ·...

Documents