Published: March 16, 2011

r 2011 American Chemical Society 2438 dx.doi.org/10.1021/ac102043p |Anal. Chem. 2011, 83, 2438–2444

ARTICLE

pubs.acs.org/ac

Screening of Protective Effect of Amifostine on Radiation-InducedStructural and Functional Variations in Rat Liver MicrosomalMembranes by FT-IR SpectroscopyGulgun Cakmak,† Faruk Zorlu,‡ Mete Severcan,§ and Feride Severcan*,†

†Department of Biological Sciences, Middle East Technical University, 06531, Ankara, Turkey‡Department of Radiation Oncology, Faculty of Medicine, Hacettepe University, 06100 Sihhiye, Ankara, Turkey§Department of Electrical and Electronic Engineering, Middle East Technical University, 06531, Ankara, Turkey

bS Supporting Information

Amifostine, which is a synthetic aminothiol compound, is theonly cytoprotective agent specifically approved by the Food

and Drug Administration (FDA) as a radioprotector.1 It is aprodrug that is dephosphorylated in the tissue by alkalinephosphatase to its active free thiol metabolite, WR-1065. In thepast, amifostine has been extensively studied in some clinicaltrials and has shown different protective activities. For example, itcan prevent or ameliorate radiation-induced xerostomia, cispla-tin-induced nephrotoxicity, anthracycline-induced cardiotoxi-city, and chemotherapy-related thrombocytopenia.2 Despitethe fact that a growing number of reports strongly supportamifostine’s clinical efficacy, the molecular effects of amifostineon the structural and functional properties of normal andirradiated tissues and membranes are largely unknown.

The deleterious effects of ionizing radiation on living cells are,for the most part, mediated by increased production of reactiveoxygen species (ROS).3 The exposure of normal biologicaltissues to such free radicals causes damage in biomoleculesresulting in several changes and finally leading to unwanted celldeath. The essential cellular target of ionizing radiation isconsidered to be DNA, with less attention being focused onbiological membranes. However, the effects of ionizing radiationon membranes deserve more interest since many physiological

processes depend on biological membranes.4 Today it is knownthat differences in lipid order, lipid dynamics, protein secondarystructure in membranes together with changes in the content ofmacromolecules disturb the kinetics and functions of ion chan-nels and also lead to onset of many diseases.5 The variations inthe ratios of biomolecules such as unsaturated to saturated lipids,and lipid to protein, are related to lipid structures such asmembrane thickness and lipid order, which are also related tomembrane permeability.5,6

Microsomes, which are subcellular particles derived from theendoplasmic reticulum upon homogenization of the tissue,contain enzymes involved in reduction and oxidation reactions,drug metabolism, cholesterol synthesis, and fatty acidmetabolism.7 In the past, it has been shown that amifostineuptake in the body is greatest in the liver, kidney, salivary glands,intestinal mucosa, and lungs.1 Therefore, the liver microsomalmembrane system is one of the best models for monitoring thedamage induced by ionizing radiation and the protecting cap-ability of amifostine in biological membranes.

Received: August 21, 2010Accepted: February 28, 2011

ABSTRACT: In this study, the protective effect of amifostine, which isthe only FDA-approved radioprotective agent, was investigated againstthe deleterious effects of ionizing radiation on rat liver microsomalmembranes at molecular level. Sprague�Dawley rats, which wereeither administered amifostine or not, were whole-body irradiated witha single dose of 800 cGy and decapitated after 24 h. The microsomalmembranes isolated from the livers of these rats were investigatedusing FT-IR spectroscopy. The results revealed that radiation caused asignificant decrease in the lipid-to-protein ratio and the degradation oflipids into smaller fragments that contain less CH2 and more carbonylesters, olefinicdCH and CH3 groups, which could be interpreted as a result of lipid peroxidation. Radiation altered the secondarystructure of proteins by inducing a decrease in the β-sheet structures and an increase in the turns and random coil structures.Moreover, a dramatic increase in lipid order and a significant decrease in the membrane dynamics were observed in the irradiatedgroup. The administration of amifostine before ionizing radiation inhibited all the radiation induced compositional, structural, andfunctional damages. In addition, these results suggest that FT-IR spectroscopy provides a novel approach to monitoring radiation-induced damage on biological membranes.

2439 dx.doi.org/10.1021/ac102043p |Anal. Chem. 2011, 83, 2438–2444

Analytical Chemistry ARTICLE

Monitoring the overall changes occurring in biological mem-branes upon exposure to ionizing radiation is a complex task. Sofar, various spectroscopic techniques, such as electron paramag-netic resonance (EPR)8 and fluorescence,9 and some biochem-ical methods10 have been employed for the study of radiation-induced damage in membranes. The biochemical techniquessuch as thiobarbituric acid reactive substances (TBARS) test andhigh-performance liquid chromatography (HPLC), which giveinformation about the concentration and content of biomole-cules, are destructive and time-consuming. Spectroscopic tech-niques such as EPR and fluorescence spectroscopy require theinsertion of either a spin label or bulky fluorophore probe intothe membrane and report changes in the surrounding localizedenvironment. These techniques were used as complementary toeach other since they give information at different time scales.Among these techniques, FT-IR spectroscopy, in recent years,has emerged as a powerful tool to probe the structure, conforma-tion, and function of lipids and proteins, simultaneously, inbiological membranes without introducing foreign perturbingprobes into the system.11�15 The observed spectral features arerelated to chemical bond vibrations of specific molecular moi-eties of individual constituents.11 These spectral parameters aresensitive to changes in the macromolecules of the biologicalsystem induced by pathological and other conditions. Therefore,FT-IR spectroscopy has recently been widely applied during theclassification and differentiation of disease states16�19 and aqua-tic toxicological states.20 It has also been employed to determineradiation-induced changes in different tissues such as food21 andrat brain.13 In the current study, ionizing radiation-inducedvariations in rat liver microsomal membrane were studied interms of macromolecular composition, structure, and lipiddynamics using FT-IR spectroscopy. In addition, the protectiveeffects of the near therapeutic doses of amifostine on radiation-induced structural and functional variations were investigatedwhich, to the best of our knowledge, have not been reportedpreviously.

’EXPERIMENTAL SECTION

Chemicals.Amifostine (WR-2721) and the chemicals used forthe isolation of microsomes were purchased from Sigma (SigmaChemical Company, Saint Louis, MO, U.S.A.). All chemicalswere obtained from the commercial source at the highest grade ofpurity available.Animals. After approval by the Ethics Committee of Hacet-

tepe University, Faculty ofMedicine, a total of 18male Sprague�Dawley rats weighing 180�220 g were housed in a room with a10:14 light/dark cycle and with free access to food and water.The animals were randomly divided into three groups. Group 1(n = 6) was used as a control group without any treatment. Therats in group 2 (n = 6) were irradiated with a single dose of 800cGy of whole body irradiation using a cobalt-60 irradiator. Therats in group 3 (n = 6) were intraperitoneously inoculated with asingle dose of 300 mg/kg body weight of amifostine 1 h prior tothe same irradiation as group 2. Because amifostine was dissolvedin isotonic saline, the animals in groups 1 and 2 were similarlyinjected with equal volumes of isotonic saline intraperitoneously.The amifostine dose was chosen in the range of therapeutic dosesused in previous studies (200�400 mg/kg body weight).22

Twenty-four hours after radiation exposure the rats wereeuthanized, and liver tissues were quickly removed and keptat �80 �C until use.

Isolation of Rat Liver Microsomal Membranes.Microsomalmembranes were isolated according to a method described bySevercan et al.14 Briefly, liver tissue was homogenized in 25 mMKH2PO4, 1.15% KCl, 5 mM EDTA, 0.2 mM PMSF, 2 mMDTT(pH 7.4) (1:4 w/v) at 4 �C. The homogenate was first centri-fuged at 16 000g for 20 min, and the resulting mitochondrialpellet was discarded. Then the supernatant was further centri-fuged at 125 000g for 60 min. The resulting pellet was suspendedin a buffer containing 50 mM Tris, 1 mM EDTA (pH 7.4) andrecentrifuged at 125 000g for 55 min. After discarding thesupernatant, the microsomal membrane-rich pellet was sus-pended with 25 mM phosphate buffer (pH 7.4) containing25% glycerol (v/v) at a volume of 0.25 mL for each gram ofprimal liver tissue. The resuspended microsomes were homo-genized manually using the Teflon�glass homogenizer andstored at �80 �C until spectroscopic studies were performed.FT-IR Spectroscopic Study and Data Analysis. In order to

remove the excessive amount of suspension buffer, the micro-somal membrane fraction was centrifuged at 14 000 rpm for10 min at 4 �C just before taking FT-IR spectra, as reportedpreviously.18 The concentrated pellet containing the liver micro-somal membranes was used in the FT-IR studies. Infrared spectrawere obtained using Perkin-Elmer Spectrum 100 FT-IR spectro-meter (Perkin-Elmer Inc., Norwalk, CT, U.S.A.) equipped with aMIR TGS detector. The sample compartment was continuouslypurged with dry air to minimize atmospheric water vaporabsorbance, which overlaps with the spectral region of interest,and carbon dioxide interference. An equal amount of sample(20 μL) was placed between CaF2 windows using a 12 μm pathlength spacer. Interferograms, both for the samples and buffer,were accumulated for 100 scans at 2 cm�1 resolution at 25 �C.Temperature regulation was performed by a Graseby Specacdigital temperature controller unit. Before data acquisition, thesamples were incubated for 10 min. Background spectra, whichwere collected under identical conditions, were used to calculatethe absorbances of the sample and buffer spectra. This procedurewas performed automatically by the appropriate software. In allof these experiments, three different independent aliquots werescanned from the same sample to check the accuracy of theabsorbance values for the same sample and the spectra werecompared. We observed that all three replicates gave identicalspectra in each tissue studied. Therefore, in each tissue sample weused the average of these three replicates to represent thespectrum of one animal. We repeated this procedure for allanimals. All these averages were then used in data analysis andstatistical analysis.The spectra of suspension buffer were subtracted from the

spectra of liver microsomal membranes to remove water absorp-tion bands using Spectrum 100 software (Perkin-Elmer) sub-traction procedure. In the subtraction process the water bandlocated around 2125 cm�1 was flattened (Supporting Informa-tion Figure S-1). The averages of the spectra belonging to thesame experimental groups, baseline correction, normalization,and the band areas were obtained by using the same software.The baseline correction and normalization process was appliedonly for visual representation of the differences. For the accuratedetermination of the mean values for the band areas, bandpositions, and bandwidth values, the original average spectrum(from three replicates) belonging to each individual of the groupswas taken into consideration. The band positions were measuredaccording to the center of mass. The bandwidth values werecalculated as the width measured at a 0.75 fraction of the

2440 dx.doi.org/10.1021/ac102043p |Anal. Chem. 2011, 83, 2438–2444

Analytical Chemistry ARTICLE

maximum height of the absorption signal in terms of wavenum-bers from raw spectra.Neural networks (NN) are used to predict the secondary

structure of proteins. Initially, NNs are trained using referenceinfrared spectra of 18 proteins whose secondary structurecharacterization are known from X-ray crystallographic analysis.The 1600�1700 cm�1 spectral region corresponding to theamide I band of the infrared spectra have been used for training.In order to improve the predictions, the data set size has beenincreased artificially to 64 by interpolating the spectra. First, thedata samples taken from the amide I band are preprocessed byamplitude normalization and then by discrete cosine transforma-tion. Bayesian regularization method was used for training theNNs. For each structure a separate NN was modeled withoptimum number of inputs and hidden neurons. During theprediction step, after applying the same preprocessing methodoptimized NNs are used to get predictions. The details of themethod can be found in Severcan et al.23

Statistical Study. Mann�Whitney U-test, which is a non-parametric test used when sample size is small, was performed onthe groups to test the significance of the differences between thecontrol and irradiated groups, as well as the amifostine treatedplus irradiated group. A p value of less than 0.05 was consideredstatistically significant.

’RESULTS AND DISCUSSION

The spectrum of rat microsomal membrane is quite complexand consists of several bands which arise from the contributionsof different functional groups. Figure 1 shows a representativeFT-IR spectrum of a control rat liver microsomal membranesample in the region of 3050�1000 cm�1. Main absorptionsobserved in the spectra are labeled in Figure 1, and detailed bandassignments are given in Table 1.

In order to observe the details of the spectral analysis, the spec-tra were investigated in two regions. Parts A and B of Figure 2show the average spectra of control, irradiated, and amifostinetreated plus irradiated rat liver microsomal membranes in the3025�2820 and 1950�1000 cm�1 regions, respectively.

The numerical comparisons of the band area ratios, frequen-cies, and bandwidths of the infrared bands are listed in Table 2.As seen from this table and Figure 2, parts A and B, there aresignificant (p < 0.05) differences between the control andirradiated group suggesting that ionizing radiation inducesremarkable changes in liver microsomal membranes. Thesefigures and the table also indicate that the amifostine treatedplus irradiated group’s spectrum is between the control andirradiated groups’ spectra and in general the values of amifostinetreated plus irradiated group are very close to those of the controlgroup. Indeed, with respect to the control, no significant varia-tion was observed in the amifostine treated group. This resultshows that amifostine has a protective effect against the ionizingradiation-induced alterations in rat liver microsomal membranes.

The signal intensity and, more accurately, the area of infraredbands arising from particular species are directly proportional tothe concentration of that species.13,20 In order to eliminate anyartifacts which may be caused by variation in experimentalconditions, e.g., sample concentration, the area ratios of somespecific infrared bands have been evaluated for quantitativecomparison between control and treated samples. Using FT-IRdata, the lipid to protein ratio can be obtained by taking the ratioof the areas of the bands arising from lipids and proteins. Theratio of the sum of the area under the CH2 asymmetric andsymmetric stretching bands to the sum of the area under theamide I and II bands was used to evaluate the lipid to protein ratioof the microsomal system. As seen from Table 2, the lipid/protein ratio decreased significantly in the irradiated liver micro-somal membranes (p < 0.05) compared to the controls. Thisdecrease could be attributed to a lower lipid and/or higherprotein content13,15 and shows that there is an alteration in the

Table 1. FT-IR Spectral Band Assignments of Rat Liver Microsomal Membrane in the Region of 3050�1000 cm�1

band no. frequency (cm�1) definition of the assignment

1 3014 olefinicdCH: unsaturated lipids (ref 14)

2 2962 CH3 asymmetric stretching: lipids and protein side chains (refs 18 and 24)

3 2925 CH2 asymmetric stretching: mainly lipids with a little contribution from proteins (refs 18 and 20)

4 2873 CH3 symmetric stretching: mainly proteins with a little contribution from lipids (refs 18 and 20)

5 2854 CH2 symmetric stretching: mainly lipids with a little contribution from proteins (ref 20)

6 1745 ester CdO stretching: lipids (ref 15)

7 1650 amide I: 80% protein CdO stretching, 10% protein N�H bending, 10% C�N stretching (ref 25)

8 1545 amide II: 60% protein N�H bending, 40% C�N stretching (ref 25)

9 1455 CH2 bending: mainly lipids with a little contribution from proteins (20)

10 1401 COO� symmetric stretching: fatty acids (ref 20)

11 1234 PO2� asymmetric stretching: phospholipids (ref 15)

12 1084 PO2� symmetric stretching: phospholipids (ref 15)

Figure 1. FT-IR spectrum of control rat liver microsomal membrane inthe 3050�1000 cm�1 region.

2441 dx.doi.org/10.1021/ac102043p |Anal. Chem. 2011, 83, 2438–2444

Analytical Chemistry ARTICLE

lipid and protein metabolism in the irradiated group. Whenamifostine was administered to the rats before radiation treat-ment, irradiation did not cause any significant alterations in thelipid/protein ratio of the liver microsomal membranes, whichindicates the protective effect of amifostine on the lipid/protein ratio.

In addition, to reveal the changes in the molecular composi-tion and structure of lipids, the ratios of some specific lipidfunctional groups (CH2 asymmetric stretching, carbonyl esterstretching, CH3 asymmetric stretching, olefinicdCH) to thetotal lipid (the sum of the CH2 asymmetric and symmetricstretching bands) were calculated. The examination of theseratios demonstrated that ionizing radiation induced importantchanges in the composition and structure of membrane lipids(Table 2). The CH2/total lipid ratio was applied to indicate thealterations in the chain length of the phospholipids, where alower ratio indicates shorter chained lipids.26 A significantdecrease in this ratio was observed in the irradiated livermicrosomal membrane (p < 0.05) compared to the control

group (Table 2). Pretreatment of amifostine suppressed thesignificant change observed in the irradiated liver microsomescompared to the control group. In order to see the carbonylstatus of the system, the carbonyl ester/total lipid ratio wascalculated. As seen from Table 2, this ratio increased significantly(p < 0.05) in the irradiated group. The values of the amifostinetreated plus irradiated group were very close to the control group,which clearly shows the protective effect of amifostine. More-over, the CH3/lipid ratio was calculated to examine the methylconcentration in the liver microsomal membrane.26 The radia-tion treatment caused an increase in the CH3/total lipid ratios inthe irradiated group compared to the control group (Table 2).Pretreatment of rats with amifostine also provided protectionagainst the increment of CH3 groups induced by ionizingradiation. To examine the unsaturation level of the system, theratio of the area of the band arising from unsaturated lipids(olefinicdCH band) to saturated lipids (CH2 asymmetric plusCH2 symmetric) was calculated. As seen from Table 2, this ratioincreased significantly in the radiation treated group (p < 0.05),whereas no significant changes were observed for amifostinetreated plus irradiated group compared to the control groupimplying that amifostine has a protective effect on the system.

Ionizing radiation is known to generate highly reactive freeradicals that can induce oxidative stress in important biologicaltargets.3 With the particular sensitivity of polyunsaturated fattyacids, biological membranes are highly vulnerable to free radicalattack.27,28 When lipids are attacked by free radicals, a lipidperoxidation chain reaction occurs. It has been found that thesedegenerative reactions lead to degradation to lipids, and theconsequences of the attack by free radicals are broken chemicalbonds, cross-linkages, and conformational changes.28 Previousreports on model membranes have shown that radiation yieldsoxidation fragments of unsaturated acyl chains.29,30 In addition, ithas been shown that peroxidation of lipids, which causes abreakdown of long chains, is usually accompanied by the forma-tion of a wide variety of degradation products including lipidaldehydes, shorter-chained lipids, and carbonyl compounds.28

Consistent with these earlier studies, in the current study thelower lipid/protein, CH2/total lipid, and the higher carbonylester/total lipid ratios observed in the irradiated group (Table 2)suggest that lipids were degraded by free radicals into smallerdegradation products which contain less CH2 and more carbonylesters. These results were supported by some previous studieswhere an increase in carbonyl groups and a degradation of acylchains were observed in vitro, using FT-IR spectroscopy afterlipid peroxidation.31 Since free radicals can also oxidize mem-brane proteins and the oxidation of proteins causes the produc-tion of some additional carbonyls, this may also contribute to theincrease in the carbonyl ester/total lipid ratio.32,33 Moreover, thebreakdown of lipid acyl chains after irradiation was confirmed byan increase in CH3/total lipid ratio which shows that thedegradation products of lipids contain more methyl groups thannormal lipids. Observations made in the present study in relationto the methyl contents are supported by other studies on theproduction of alkyl radicals during lipid peroxidation induced bydifferent agents.34,35 The increase in the unsaturated/saturatedlipid ratio may be due to higher lipid peroxidation in theirradiated membrane ultimately inducing an increase in theolefinic content originating mainly from lipid peroxidation end-products such as malondialdehyde, lipid aldehydes, and alkylradicals.14,36 Consequently, the lower lipid/protein and CH2/total lipid ratio and the higher carbonyl ester/total lipid, CH3/

Figure 2. Average baseline-corrected infrared spectra of control, irra-diated, and amifostine treated plus irradiated rat liver microsomalmembranes (A) in the 3025�2820 cm�1 region and (B) in the 1950�1000 cm�1 region. The spectra were normalized with respect to theamide I band (A) and to the CH2 asymmetric stretching band (B).

2442 dx.doi.org/10.1021/ac102043p |Anal. Chem. 2011, 83, 2438–2444

Analytical Chemistry ARTICLE

total lipid, and unsaturated lipid/saturated lipid ratios suggestthat lipids were oxidized by ionizing radiation, and as a result ofthis oxidation some degradation products which contain lessCH2 group and more carbonyl esters, CH3, and unsaturated fattyacids were produced. As seen from Table 2, amifostine was verysuccessful in protecting the microsomal membrane lipids fromperoxidation damage induced by ionizing radiation.

FT-IR spectroscopy offers unique possibilities for the simul-taneous study of proteins together with lipids in biologicalmembranes and tissues.6,12,20 To evaluate the changes in proteincomposition and structure, the amide I/amide II ratio wascalculated. We observed a significant decrease in the amide I/amide II ratio of the irradiated group (p < 0.05) in contrast to theamifostine treated plus irradiated group in comparison to thecontrol group (Table 2). As the amide I and amide II profilesdepend on the protein structural composition, this reductionsuggests that there are some alterations in the structures ofmembrane proteins.37�39 Consistent with our result, Dogan et al.observed a reduction in the amide I/amide II ratio after γ-irradiation in hazelnut tissue, and they suggested that thisreduction was associated with changes in the protein structure.21

This explanation can be reliable since amide I band arisespredominantly from the CdO stretching vibration of the amidegroups, whereas the amide II band arises predominantly from theN�H bending vibration.25 Moreover, in recent medical researchit has been shown that, compared with the normal tissue, thediseased tissue shows a change in protein secondary structureand a reduced ratio of amide I to amide II ratio.40 In addition, asignificant shifting in the frequency to higher values of the amideI band (p < 0.05) and a significant broadening in the bandwidthof this band (p < 0.05) were observed in the irradiated group (p <0.05) in comparison to the control group (Table 2). Thisbroadening and shifting observed in the amide I band indicatesthat the protein conformation was affected from ionizingradiation.20 In a previous study, protein oxidation was found tobroaden the amide I band.41 As noted above, protein oxidationresults in the production of some additional carbonyls on someamino acid residues.32 It has been suggested that some of these

carbonyls reside adjacent to amines and they lead to a spectro-scopic absorption and a broadening in the amide I band.32,41

Thus, the decrease in the amide I/amide II ratio, and thebroadening and shifting of the amide I band to higher values,could be interpreted as the result of the altering protein structureand conformation of irradiated liver microsomal membranes.

In order to better estimate the alterations on the proteinsecondary structure, the amide I band region, corresponding toabsorption values between 1600 and 1700 cm�1, was analyzedusing neural network predictions based on the FT-IR data. FT-IRspectroscopy is an excellent method for determining proteinsecondary structure. Neural networks based on the amide I bandoffer a new computational approach that has served as a reliablealternative method for predicting protein secondary structure inrecent years in solutions23,42 and by our group in biologicaltissues.43�46 The results are presented in Table 3, which revealedthat the β-sheet structures decreased significantly in the irra-diated group (p < 0.05) compared to the control group. More-over, the turns and random coil structures increased in theirradiated group (p < 0.05) compared to the control group.The increase in the random coil structure indicates that ionizingradiation caused protein denaturation in the system.21 Ourresults are in agreement with earlier studies that reported anincrease in the random coil structure.21,47 Moreover, Abu et al.reported protein denaturation due to γ-irradiation, which alsosupports our findings.48 A metabolic disruption of some aminoacids could occur after radiation treatment in liver microsomesand could be responsible for the observed changes in theproteins. It is known that free radicals lead to protein degradationby increasing accessibility of protein bonds and/or aggregation.49

In a previous study, it has been reported that radiation causes analteration in the conformation of integral membrane proteinsand that radiation-induced hydroxyl radicals lead to denaturationof membrane proteins.50 On the other hand, a relationshipbetween lipid peroxidation and damage to protein synthesishas been previously reported in liver slices from rats.51 As seenfrom Table 3, there are no significant differences in the proteinsecondary structure profile of the amifostine treated plus

Table 2. Changes in the Band Area Ratio, Frequency, and Bandwidth Values of Various Functional Groups in Control, Irradiated,and Amifostine Treated Plus Irradiated Liver Microsomal Membranes

functional group control radiation P valuea radiation plus amifostine P valueb

BandArea Ratio

lipid/protein 0.064( 0.004 0.054 ( 0.007 0.035* 0.062( 0.004 0.575

CH2/lipid 0.874( 0.05 0.816( 0.007 0.030* 0.831 ( 0.009 0.173

carbonyl ester/lipid 0.102( 0.005 0.115( 0.006 0.013* 0.108( 0.008 0.128

CH3/lipid 0.129( 0.015 0.156( 0.015 0.030* 0.124( 0.011 0.378

unsaturation/saturation 0.042( 0.005 0.049( 0.003 0.030* 0.046( 0.006 0.229

amide I/amide II 1.847( 0.289 1.522 ( 0.144 0.022* 1.615( 0.139 0.065

Frequency

CH2 asym. str. 2924.85 ( 0.16 2923.94( 0.22 0.005** 2924.53( 0.09 0.200

amide I 1649.90( 0.92 1651.11( 0.70 0.030* 1650.64( 0.55 0.173

Bandwidth

CH2 asym. str. 12.338( 0.23 12.038( 0.12 0.037* 12.20( 0.11 0.378

amide I 27.972( 0.68 29.514( 1.25 0.035* 28.56( 0.85 0.229a P value: p values are from the comparison of the control and radiation groups. b P value: p values are from the comparison of the control and radiationplus amifostine groups.

2443 dx.doi.org/10.1021/ac102043p |Anal. Chem. 2011, 83, 2438–2444

Analytical Chemistry ARTICLE

irradiated group compared to the control group supporting theprotective effect of amifostine on protein secondary structure.

The position of the asymmetric and symmetric CH2 stretchingbands provides information about the lipid acyl chain flexibility(order/disorder state of lipids). For example, the higher thefrequency, the higher the acyl chain flexibility, which implies lipiddisordering.11,12,36 The frequency of the CH2 asymmetricstretching band shifted significantly to lower values after radia-tion treatment (p < 0.05) in comparison to the control. Thisshifting to lower values indicates that lipid order increases andacyl chain flexibility decreases in the irradiated liver microsomalmembranes.12,20 Bandwidth of the same band gives informationabout the dynamics of the system.20 As seen from Table 2, thereis a significant reduction in the bandwidth of the CH2 asymmetricstretching band of the irradiated group in comparison to thecontrol group, which implies a decrease in the lipid dynamics ofthe radiation-treated liver microsomal membranes. Thus, ioniz-ing radiation caused an increase in the lipid order and a decreasein the membrane fluidity in the irradiated microsomal mem-branes. Liver microsomes contain cholesterol and a great varietyof phospholipids at high concentrations. The fatty acid composi-tion of rat liver microsomes is so highly unsaturated that they arevery fluid at physiological temperatures.52 Previously, the loss offreedom of motion in membranes after oxidative stress usingbiological and liposomal membranes was reported,53�55 andlipid peroxidative stress was known to decrease membranefluidity in microsomes and other cellular membranes.3,56 Animportant factor affecting the membrane structure and dynamicsis the amount of proteins and lipids in the membranes.6 Twoimportant reasons could be proposed as causal relationships forthe loss of membrane fluidity during ionizing radiation treat-ment. First, radiation may cause cholesterogenesis and due to theincreased amount of cholesterol a decrease in membrane fluiditymay be observed; second, the formation of cross-linking amongthe lipid�lipid and protein�lipid moieties as a result of oxidativestress may limit the motion.54,55,57 As seen from Table 2, nosignificant changes were observed in the frequency and band-width of the CH2 asymmetric stretching band of the amifostinetreated plus irradiated group in comparison to the control group,again indicating the restoring effect of amifostine. This resultshows that stabilizing of cell membranes may be anotherimportant function by which amifostine reduces oxidative da-mage to cells. Alterations in membrane fluidity have importantconsequences in terms of cellular function.5,6 Thus, the ability ofamifostine to maintain membrane fluidity may further contributeto the protective actions of this molecule.

In all cases, the damage induced by ionizing radiation onmembrane lipids and proteins was suppressed by amifostineadministration. The inhibitory effects of amifostine against lipidand protein damage could be attributed to its antioxidant

properties. WR-1065, the active form of amifostine, must bepresent at the time of radiation exposure.1 In the cell, WR-1065can provide protection by several mechanisms. It can act as apotent scavenger of oxygen free radicals derived from radiationtherapy.58 Another mechanism of cytoprotection involves hydro-gen donation to repair free radicals in target molecules.58 WR-1065may also react directly with oxygen and thus protect the cellby creating local hypoxia at the target.58 Other mechanismsproposed for the radioprotective effect of amifostine are targetstabilization by binding to DNA and the protection of keysulfhydryl enzymes through formation of protein�amino thiolmixed disulfides.59

’CONCLUSIONS

In conclusion, the results of the present study revealed thationizing radiation induced significant alterations in the composi-tion and structure of microsomal lipids, dramatic increases inlipid order, and significant decreases in membrane dynamics.Moreover, the composition and secondary structure of proteinswere altered by ionizing radiation. Amifostine administration tothe rats prior to whole body irradiation was effective in prevent-ing the radiation-induced damages on membrane lipids andproteins, likely through a free radical scavenger activity. Wesuggest that maintaining membrane fluidity despite the rigidityeffect of ionizing radiation may be another cytoprotectivemechanism of amifostine. In addition, the current study indicatedthat FT-IR spectroscopy provides a rapid and sensitive monitor-ing of ionizing radiation-induced damage in biological mem-branes and FT-IR parameters employed in this study can be usedas biological indicators of radiation-induced membrane damage.

’ASSOCIATED CONTENT

bS Supporting Information. Additional information asnoted in text. This material is available free of charge via theInternet at http://pubs.acs.org.

’AUTHOR INFORMATION

Corresponding Author*Phone: þ90-312-210 51 66. Fax: þ90-312-210 79 76. E-mail:feride@metu.edu.tr.

’ACKNOWLEDGMENT

This work was supported by the Scientific and TechnicalResearch Council of Turkey (TUBITAK), SBAG-2939 ResearchFund, and by theMETU-Research Fund, BAP-2006-07-02-0001.

Table 3. Neural Network Predictions Based on FT-IRData in the 1600�1700 cm�1 Spectral Region for the Changes in SecondaryStructure between the Control, Irradiated, and Amifostine Treated Plus Irradiated Liver Microsomal Membranes

functional group control radiation P valuea radiation plus amifostine P valueb

R-helix 40.94( 1.62 39.40( 2.96 0.117 41.15( 0.94 0.855

β-sheet 26.30( 1.24 24.22( 0.82 0.028* 25.50( 1.55 0.361

turns 20.50( 0.47 21.88( 0.39 0.012* 20.46( 0.48 0.927

random coil 12.26( 1.11 14.82( 0.13 0.012* 13.05( 0.51 0.235a P value: p values are from the comparison of the control and radiation groups. b P value: p values are from the comparison of the control and radiationplus amifostine groups.

2444 dx.doi.org/10.1021/ac102043p |Anal. Chem. 2011, 83, 2438–2444

Analytical Chemistry ARTICLE

’REFERENCES

(1) Grdina, D. J.; Murley, J. S.; Kataoka, Y. Oncology 2002, 63, 2–10.(2) Marzatico, F.; Porta, C.; Moroni, M.; Bertorelli, L.; Borasio, E.;

Finotti, N.; Pansarasa, O.; Castagna, L. Cancer Chemother. Pharmacol.2000, 45, 172–176.(3) Karbownik, M.; Reiter, R. J. Proc. Soc. Exp. Biol. Med. 2000,

225, 9–22.(4) Hidalgo, A. A.; Tabak, M.; Oliveira, O. N., Jr. Chem. Phys. Lipids

2005, 134, 97–108.(5) Awayda, M. S.; Shao, W.; Guo, F.; Zeidel, M.; Hill, W. G. J. Gen.

Physiol. 2004, 123, 709–727.(6) Szalontai, B.; Nishiyama, Y.; Gombos, Z.; Murata, N. Biochim.

Biophys. Acta 2000, 1509, 409–419.(7) Garg, M. L.; Mcmurchi, E. J; Sabine, J. R. Biochem. Int. 1985,

11, 677–686.(8) Bartosz, G.; Schon, W.; Kraft, G.; Gartner, H. Radiat. Environ.

Biophys. 1992, 31, 117–121.(9) Kolling, A.; Maldonado, C.; Ojeda, F.; Diehl, H. A. Radiat.

Environ. Biophys. 1994, 33, 303–313.(10) Pribrush, A.; Agam, G.; Yermiahu, T.; Dvilansky, A.; Meyer-

stein, D.; Meyerstein, N. Free Radical Res. 1994, 21, 135–146.(11) Lewis, E. N.; Levin, I. W.; Steer, C. J. Biochim. Biophys. Acta

1989, 986, 161–166.(12) Toyran, N.; Zorlu, F.; D€onmez, G.; O�ge, K.; Severcan, F. Eur.

Biophys. J. 2004, 33, 549–554.(13) Toyran, N.; Zorlu, F.; Severcan, F. Int. J. Radiat. Biol. 2005,

81, 911–918.(14) Severcan, F.; Gorgulu, G.; Gorgulu, S. T.; Guray, T. Anal.

Biochem. 2005, 339, 36–40.(15) Akkas, S. B.; Inci, S.; Zorlu, F.; Severcan, F. J. Mol. Struct. 2007,

834, 207–215.(16) Kneipp, J.; Beekes, M.; Lasch, P.; Naumann, D. J. Neurosci.

2002, 22, 2989–2997.(17) Levin, I. W.; Bhargava, R. Annu. Rev. Phys. Chem. 2005,

56, 429–474.(18) Severcan, F.; Bozkurt, O.; Gurbanov, R.; Gorgulu, G.

J. Biophotonics 2010, 3, 621–631.(19) Leskovjan, A. C.; Kretlow, A.; Miller, L. M. Anal. Chem. 2010,

82, 2711–2716.(20) Cakmak, G.; Togan, I.; Severcan, F. Aquat. Toxicol. 2006,

77, 53–63.(21) Dogan, A.; Siyakus, G.; Severcan, F. Food Chem. 2007,

100, 1106–1114.(22) Jirtle, R. L.; Pierce, L. J.; Crocker, I. R.; Strom, S. C. Radiother.

Oncol. 1985, 4, 231–237.(23) Severcan, M.; Severcan, F.; Haris, I. P. Anal. Biochem. 2004,

332, 238–244.(24) Szalontai, B. PMC Biophys. 2009, 2, 1–17.(25) Haris, P. I.; Severcan, F. J. Mol. Catal. B: Enzym 1999,

7, 207–221.(26) Antoine, K. M.; Mortazavi, S.; Miller, A. D.; Miller, L. M.

J. Forensic Sci. 2010, 55, 513–518.(27) Reiter, R. J. FASEB J. 1995, 9, 526–533.(28) Zwart, L. L.; Meerman, J. H. N.; Commandeur, J. N. M.;

Vermeulen, N. P. E. Free Radical Biol. Med. 1999, 26, 202–226.(29) Tait, D.; Bors, W. Int. J. Radiat. Biol. 1985, 47, 497–508.(30) Yukawa, O.; Miyahara, M.; Shiraishi, N.; Nakazawa, T. Int. J.

Radiat. Biol. 1985, 48, 107–115.(31) Lamba, O. P.; Lal, S.; Yappert, M. C.; Lou, M. F.; Borchman, D.

Biochim. Biophys. Acta 1991, 1081, 181–187.(32) Stadtman, E. R. Free Radical Biol. Med. 1990, 9, 315–325.(33) Manda, K.; Ueno, M.; Moritake, T.; Anzai, K. Cell Biol. Toxicol.

2007, 23, 129–137.(34) Sonia, H.; Maria, B. K.; Ronald, P. M.; Ohara, A. Chem. Res.

Toxicol. 2000, 13, 1056–1064.(35) Koshiishi, I.; Tsuchida, K.; Takajo, T.; Komatsu, M. J. Lipid Res.

2005, 46, 2506–2513.

(36) Liu, K. Z.; Bose, R.; Mantsch, H. H. Vib. Spectrosc. 2002,28, 131–136.

(37) Yu, P.; Kevin, D.; Dasen, L. J. Agric. Food Chem. 2008,56, 3417–3426.

(38) Ishida, K. P.; Griffiths, P. R. Appl. Spectrosc. 1993, 47, 584–589.(39) Schmidt, M.; Wolfram, T.; Rumpler, M.; Tripp, C. P.; Grunze,

M. Biointerphases 2007, 2, 1–5.(40) Szczerbowska-Boruchowska, M.; Dumas, P.; Kastyak, M. Z.;

Chwiej, J.; Lankosz, M.; Adamek, D.; Krygowska-Wajs, A.Arch. Biochem.Biophys. 2007, 459, 241–248.

(41) Signorini, C.; Ferrali, M.; Ciccoli, L.; Sugherini, L.; Magnani, A.;Comporti, M. FEBS Lett. 1995, 362, 165–170.

(42) Hering, J. A.; Innocent, P. R.; Haris, P. I. Proteomics 2004,4, 2310–2319.

(43) Garip, S.; Yapici, E.; Ozek, N. S.; Severcan, M.; Severcan, F.Analyst 2010, 135, 3233–3241.

(44) Bozkurt, O.; Severcan, M.; Severcan, F. Analyst 2010,135, 3110–3119.

(45) Akkas, S. B.; Severcan, M.; Yilmaz, O.; Severcan, F. Food Chem.2007, 105, 1281–1288.

(46) Toyran, N.; Severcan, F.; Severcan, M.; Turan, B. Spectrosc.—Int. J. 2007, 215, 269–278.

(47) Lee, S.; Song, K. B. Food Chem. 2003, 82, 521–526.(48) Abu, J. O.; Muller, K.; Duodu, K. G.; Minnaar, A. Food Chem.

2006, 95, 138–147.(49) Davies, K. J. J. Biol. Chem. 1987, 262, 9895–9901.(50) Perromat, A.; Melin, A. M.; Lorin, C.; Deleris, G. Biopolymers

2003, 72, 207–216.(51) Fraga, C. G.; Zamora, R.; Tappel, A. L. Arch. Biochem. Biophys.

1989, 270, 84–91.(52) Catala, A. Mol. Cell. Biochem. 1993, 120, 89–94.(53) Dobretsov, G. E.; Borschevskaya, T. A.; Petrov, V. A.;

Vladimirov, Y. A. FEBS Lett. 1977, 84, 125–128.(54) Curtis, M. T.; Gilfor, D.; Farber, J. L. Arch. Biochem. Biophys.

1984, 35, 644–649.(55) Chen, J. J.; Yu, B. P. Free Radical Biol. Med. 1994, 17, 411–418.(56) Garcia, J. J.; Reiter, R. J.; Guerrero, J. M.; Escames, G.; Yu, B. P.;

Oh, C. S.; Munoz-Hoyos, A. FEBS Lett. 1997, 408, 297–300.(57) Garcia, J. J.; Reiter, R. J.; Pi�e, J.; Ortiz, G. G.; Cabrera, J. J.; S�ainz,

R. M.; Acu~na-Castroviejo, D. J. Bioenerg. Biomembr. 1999, 31, 609–616.(58) Andreassen, C. N.; Grau, C.; Lindegaard, J. C. Semin. Radiat.

Oncol. 2003, 13, 62–72.(59) Hospers, G. A.; Eisenhauer, E. A.; de Vries, E. G. Br. J. Cancer

1999, 80, 629–638.