PUBLICATIONS

� Richa Trivedi, Ahmad Raza Khan, Poonam Rana, Seenu Haridas, BS Hemanth

Kumar, Kailash Manda, Ram KS Rathore, Rajendra P Tripathi, Subash Khushu.

Radiation- Induced Early Changes in the Brain and Behavior: Serial Diffusion Tensor

Imaging and Behavioral Evaluation after Graded Dose of Radiation. J. Neurosc. Res.

(Accepted)

� A.R. Khan, P. Rana, R. Tyagi, M. M. Devi, S. Chaturvedi, S. Javed, and S.

Khushu. Identification of pathophysiological perturbations and metabolic

phenotypes based on urinary NMR spectral profiling after graded radiation

exposure in mice. (Submitted)

� Poonam Rana, Ahmad Raza Khan, Shipli Modi, B. S. Hemanth Kumar, Salim

Javed, Rajendra Prasad Tripathi and Subash Khushu.

Brain Metabolism Impairment

after Whole-body Irradiation in Mice: An in-vivo 1H Magnetic Resonance

Spectroscopy Study. (Submitted after revision)

� Ahmad Raza Khan, Poonam Rana, Ritu Tyagi, Indracanti Prem Kumar, M. Memita

Devi, Salim Javed, Rajendra P. Tripathi, Subash Khushu. NMR spectroscopy based

metabolic profiling of urine and serum for investigation of physiological perturbations

during radiation sickness. Metabolomics (2011). DOI: 10.1007/s 1306-011-02774.

� Ahmad Raza Khan, Poonam Rana, M Memita Devi, Shubhra Chaturvedi, Salim

Javed, Rajendra P Tripathi, Subash Khushu. Nuclear Magnetic Resonance

Spectroscopic-based metabonomic investigation of the biochemical effects in serum

of γ irradiated mice. Intl. J. Rad .Biol. (2011), 87 (1); 91-97.

� Ritu Tyagi, Poonam Rana, Ahmad Raza Khan, Deepak Bhatnagar, M Memita

Devi, Shubhra Chaturvedi, Rajendra P. Tripathi, Subash Khushu. Study of acute

biochemical effects of thallium toxicity in mouse urine by NMR spectroscopy.

J.Appl. Toxicol. (2010) doi:10.1002/jat.1617

� Ritu Tyagi, Poonam Rana, Mamta Gupta, Ahmad Raza Khan, Deepak Bhatnagar,

P J S Bhalla, Shubhra Chaturvedi Rajendra P Tripathi, Subash Khushu. Differential

Biochemical Response of Rat Kidney towards Low and High Doses of NiCl2 as

Revealed by NMR Spectroscopy. J.Appl. Toxicol. (2011).doi:10.1002/jat.1730

� Ritu Tyagi, Poonam Rana, Mamta Gupta, Ahmad Raza Khan, M Memita

Devi,Deepak Bhatnagar, Raja Roy, Rajendra P Tripathi, Subash Khushu. Urinary

Metabolomic Phenotyping of Nickel induced Acute Toxicity in Rat: A NMR

Spectroscopy approach. Metabolomics (Accepted).

PAPER PRESENTED IN CONFERENCES

� Mamta Gupta, Richa Trivedi, Poonam Rana, Ahmad Raza Khan, Ravi Soni, Subash

Khushu. Differential Effects of Fractionated and Single Radiation Dose on Diffusion

Tensor Imaging in Mice Brain Proc. Intl. Soc. Mag. Reson. Med, Melbourne,

Australia, (2012).

� Mamta Gupta, Poonam Rana, Richa Trivedi, Ahmad Raza Khan, Ravi Soni1,

Subash Khushu. A Comparative Evaluation of Brain Metabolites following Whole

Body and Cranial Irradiation: A Prospective 1H MRS Study. Proc. Intl. Soc. Mag.

Reson. Med, Melbourne, Australia, (2012).

� Ahmad Raza Khan, Poonam Rana, Shubhra Chaturvedi and Subash Khushu. Acute

effect of gamma irradiation in mice by NMR based metabolic profiling of urine.

Proc. Intl. Soc. Mag. Reson. Med, Stockholm, Sweden,(2010)

� Poonam Rana, Ahmad Raza Khan, Shubhra Chaturvedi and Subash Khushu. NMR

spectroscopy based study of physiological perturbations during recurrence of

symptoms in radiation sickness. Proc. Intl. Soc. Mag. Reson. Med, Stockholm,

Sweden, (2010)

� Ritu Tyagi, Poonam Rana, Ahmad Raza Khan, Memita Devi, Shubhra Chaturvedi

and Subash Khushu. Toxicological effect of thallium in mice by NMR-based

metabolic profiling of urine. Proc. Intl. Soc. Mag. Reson. Med, Stockholm, Sweden,

(2010

� Ahmad Raza Khan, Poonam Rana, M. Memita Devi, Shubhra Chaturvedi and

Subash Khushu. NMR spectroscopy based study of physiological perturbations in

radiation sickness Poster presented at 16th

NMRS Meeting, Held at CBMR, SGPGI,

Lucknow (2010).

� Ahmad Raza Khan, Poonam Rana, Sahil Moza, Shubhra Chaturvedi and Subash

Khushu. NMR spectroscopic based metabonomic investigation on the biochemical

effects induced by γ radiation exposure in mouse serum Proc. Intl. Soc. Mag.

Reson. Med.Honolulu, Hawaii,(2009).

� Poonam Rana, Ahmad Raza Khan, Poonam Singh, Shubhra Chaturvedi, Rajendra

Prasad Tripathi, Subash Khushu. High Resolution MR Spectroscopy Reveals

Radiation Induced Metabolic Changes in Mouse Liver Tissues. Proc. Intl. Soc. Mag.

Reson. Med.Honolulu, Hawaii,(2009).

� Poonam Rana, Ritika Agarwal, Ahmad Raza Khan, Shubhra Chaturvedi, Rajendra

Prasad Tripathi, Subash Khushu. Exploration of the Effect of Whole Body Ionising

Radiation Exposure on the Metabolism of Renal Tissue in Mice Using High

Resolution 1H NMR Spectroscopy. Proc. Intl. Soc. Mag. Reson. Med.Honolulu,

Hawaii,(2009).

� Poonam Rana, Ahmad Raza Khan, Ritika , Poonam Singh, Shubhra Chaturvedi,

and Subash Khushu. .Exploration of the Effect of Whole Body Ionizing

Radiation Exposure on the Renal and Hepatic Tissue Metabolism in Mice Using

High Resolution H NMR Spectroscopy. Poster presented at 15th

NMRS Meeting,

Held at IICT, Hyderabad (2009).

� Poonam Rana, Ahmad Raza Khan, Sahil Moza, Shubhra Chaturvedi and

Subash Khushu. Metabolic Profiling of Serum from Mice under Radiation Stress

using High Resolution NMR Spectroscopy. IISc Centenary Symposium on

Future Directions in NMR October 19-22, 2008.

� P. Rana, A. Varshney, G. Shreepathy, Ahmad Raza Khan, R.K. Marwah and S.

Khushu (2008). Muscle Bioenergetics in subclinical hypothroidism: A 31P MRS

study. Poster presented at 14th

NMRS Meeting, Held at INMAS, Delhi. 16th

-19th

Jan

2008.

Nuclear magnetic resonance spectroscopy-based metabonomicinvestigation of biochemical effects in serum of g-irradiated mice

AHMAD RAZA KHAN1, POONAM RANA1, M. MEMITA DEVI1,

SHUBHRA CHATURVEDI2, SALIM JAVED3, RAJENDRA P. TRIPATHI1, &

SUBASH KHUSHU1

1NMR Research Centre, and 2Division of Cyclotron & Radiopharmaceutical Sciences, Institute of Nuclear Medicine and Allied

Sciences (INMAS), India, and 3Department of Biochemistry, Jamia Hamdard, Delhi, India

(Received 13 January 2010; Revised 22 July 2010; Accepted 29 July 2010)

AbstractPurpose: Radiation exposure induces change in many biological compounds. It is important to assess the physiological andbiochemical response to an absorbed dose of ionising radiation due to intentional or accidental event and to predict medicalconsequences for medical management. In the present study, nuclear magnetic resonance (NMR) spectroscopy-basedmetabolic profiling was used in mice serum for identification of radiation-induced changes at metabolite level.Materials and methods: Mice were irradiated with 3, 5 and 8 Gray of g-radiation dose and serum samples collected at day 1,3 and 5 post irradiation were analysed by proton nuclear magnetic resonance (1H NMR) spectroscopy. 1H NMR spectra ofserum were analysed by pattern recognition using principal component analysis.Results: Irradiated mice serum showed distinct metabonomic phenotypes and revealed dose- and time-dependent clustering ofirradiated groups. 1H NMR spectral analysis exhibited increased lactate, amino acids, choline and lipid signals as well asdecreased glucose signals. These findings indicate radiation-induced disturbed energy, lipid and protein metabolism.Conclusions: The information obtained from this study reflects multiple physiological dysfunctions. The study promises theapplication of NMR-based metabonomics in the field of radiobiology, for development of metabolic-based markers forscreening of risk populations and medical management in these cases.

Keywords: radiation, NMR spectroscopy, serum, metabonomics

Abbreviations: NMR, nuclear magnetic resonance; FID, free induction decay; PCA, principal components analysis;PR, pattern recognition; BCA, branched chain amino acid; PC, principal component; TSP, 3-(Trimethylsilyl)propionic-2, 2, 3, 3-d4 acid sodium salt; Gy, gray; LDL, low density lipoprotein; VLDL, very low density lipoprotein;ROS, reactive oxygen species.

Introduction

The study of normal tissue response and injury after

exposure to moderate dose of ionising radiation is of

great importance to patients with cancer, and people

potentially subjected to military, nuclear power

industry and radiological terrorism (Coleman et al.

2003). The resulting radiation injuries would be due

to a series of molecular, cellular, tissue and whole

organal processes. During the last decade, research

in the field of radiation sciences has been more

focused on the development of a multiparametric

approach for studying gene, protein and other

biochemical expressions after radiation exposure

(Bertho et al. 2008). To develop a multiparametric

approach, inputs from all relevant fields using high

throughput technology are required at cellular,

biochemical and molecular levels since damage from

ionising radiation occurs at cellular and systemic

levels.

Since the last decade, the metabonomics approach

is becoming increasingly popular as the source of

biomedical data. Mass spectrometry and proton

nuclear magnetic resonance (1H NMR) spectroscopy

are the two key experimental methods applied in this

area of global biochemistry. In particular, 1H NMR

spectroscopy has the advantage of efficiently obtain-

ing information on large number of metabolites in

Correspondence: Dr Subash Khushu, NMR Research Centre, INMAS, DRDO, Lucknow Road, Timarpur, Delhi, India. Tel: þ91 11 23905313.

Fax: þ91 11 23919509. E-mail: skhushu@yahoo.com

Int. J. Radiat. Biol., 2010, Early Online, pp. 1–7

ISSN 0955-3002 print/ISSN 1362-3095 online � 2011 Informa UK, Ltd.

DOI: 10.3109/09553002.2011.518211

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y In

stitu

te o

f N

ucle

ar M

edic

ine

& A

llied

Sci

enes

on

12/0

9/10

For

pers

onal

use

onl

y.

biofluids such as serum, in vitro, ex vivo and in vivo

in various tissues that could serve as snapshots of

physiological states (Ala-Korpela 2008). Proton

nuclear magnetic resonance (NMR) spectrum en-

ables the simultaneous identification and monitoring

of a wide range of low molecular weight endogenous

metabolites, thus providing a biochemical fingerprint

of an organism (Nicholson and Wilson 1989). It is

now well established that this profile gets altered in

disease or toxic process and a metabonomic

approach can be successfully applied to various

toxicological, disease evaluation and risk assessment

studies (Holmes et al. 2000, Lindon et al. 2004,

Serkova and Niemann 2006, Ala-Korpela 2008).

Several biological compounds are altered in body

tissues after radiation exposure (Walden and Farza-

neh 1990). However, little is known on how the body

responds to ionising radiation at a metabolite level,

which may represent the most incisive indicator of

cellular physiology. Recently, a few metabolomic

studies based on ultra performance liquid chromato-

graphy (UPLC), and gas chromatography-mass

spectrometry (GC-MS) have reported several meta-

bolites of energy, purine and pyrimidine metabolism

as urinary biomarkers in rodents with radiation

exposure (Patterson et al. 2008, Tyburski et al.

2008, Lanz et al. 2009). It is necessary to explore

metabolic markers of ionising radiation exposure

based on a physiological and biochemical response

that can be used for mass screening in the events of

radiological incidents. Analysis of biofluids such as

serum and urine is the simplest and minimally

invasive approach to work at a mass screening level.

The present study was undertaken to investigate the

biochemical and physiological perturbations after

ionising radiation exposure in mouse serum using

metabonomic approach based on high resolution

NMR spectroscopy.

Materials and methods

Animal handling, radiation exposure and sample

collection

Strain ‘A’, male mice of 10 weeks of age obtained

from experimental animal facility of the institute

were placed in polypropylene cages. The mice

received tap water, were fed chow ad libitum and

were housed under a standard 12 h light/12 h dark

cycle and at controlled temperature of 25+ 28C and

relative humidity of 45–65%. All experimental

protocols and animal handling practices adopted in

this study were approved by institutional animal

ethical committee. Three groups with an equal

number of animals in each group (n¼ 18) were

exposed to gamma radiation doses of 3 (group I), 5

(group II) and 8 Gray (Gy) (group III), respectively,

and controls (n¼ 9, group IV) were sham irradiated.

The mice were irradiated through gamma irradiation

facility 60Co (Gammacell-220, Atomic Energy of

Canada Ltd (AECL), Ottawa, Ontario, Canada).

During the course of experimentation, the dose rate

of the source was 0.22 Gy/min. Irradiation was

performed on mice kept in pi-cage which was placed

in irradiation chamber. Air supply was established

through air pump in irradiation chamber to avoid

hypoxic condition. Approximately 400 ml blood

sample was collected between 09:00 and 10:00 h

through retro-orbital plexus route from controls and

irradiated animals. In the irradiated group, blood was

collected at three time points (day 1, 3 and 5) post

irradiation from one third of animals present in group

at each time point (n¼ 6, animals for each dose level

at each time point). Blood sample was allowed to clot

for 30 min and serum was separated by centrifuga-

tion at 2665 g for 10 min. All serum samples were

stored at7808C until NMR measurement.

Sample preparation and NMR spectroscopy

NMR solvents, trimethylsilyl-2,2,3,3-tetradeutero-

propionic acid (TSP) and deuterium oxide (D2O)

were obtained from Sigma-Aldrich (St Louis, MO,

USA). Serum sample of 200 ml was mixed with

400 ml of D2O and transferred to 5 mm NMR tubes

containing 1 mM TSP closed capillary (Wilmad, SP

Industries, Buena, NJ, USA) as reference. All NMR

spectra were acquired at 400.13 MHz, Bruker-AV

spectrometer (Bruker, Rheinstetten, Karlsruhe Ger-

many) at 298 K. Water suppressed Carr-Purcell-

Meiboom-Gill (CPMG) spin echo pulse sequence

(RD-908-(t71808-t) n-acquire) with a total spin

echo (2 nt) of 200 milliseconds was used to attenuate

broad signals from proteins and lipoproteins. Here

RD represents relaxation delay of 2 sec during which

water resonance is selectively irradiated, 908 repre-

sents a 908 radio frequency (RF) pulse, tau (t) is spin

echo delay, 1808 is 1808 RF pulse and n represents

number of loops. Typically 64 free induction decays

(FID) were collected into 32 K data points over a

spectral width of 8223.68 Hz with an acquisition

time of 3.98 sec. Each FID was weighted by an

exponential function with a 0.3-Hz line broadening

factor prior to Fourier transformation. 1H NMR

chemical shifts (d) in the spectra were referenced to

TSP at d 0.0 ppm and peaks identified on the basis of

chemical shifts were assigned according to previously

reported literature (Lindon et al. 1999).

Data reduction and pattern recognition (PR) analysis of1H NMR spectra

All the spectra were phase and baseline-corrected

using TOPSPIN (Bruker, Rheinstetten, Karlsruhe,

2 A. R. Khan et al.

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y In

stitu

te o

f N

ucle

ar M

edic

ine

& A

llied

Sci

enes

on

12/0

9/10

For

pers

onal

use

onl

y.

Germany). The spectra over the range (d 0.2–10.0

ppm) were reduced to 245 regions, each 0.04 ppm

wide and signal intensity in each region was

integrated. The region between d 4.5 and d 5.0

ppm was removed prior to any statistical analysis in

order to remove any spurious effects of variability in

the suppression of the water resonance. Following

removal of this region, each integral region was

divided by the sum of all integral regions for data

normalisation. PR analysis was done by carrying out

principal components analysis (PCA) on normalised

data. PCA is a well known statistical procedure for

data dimensionality reduction. It allows the expres-

sion of most of the variance within a data set in a

small number of newly transformed variables known

as principal component (PC). Each PC is a linear

combination of original variables whereby each

successive PC explains the maximum amount of

variance possible that is not accounted for by the

previous PC. Hence, each PC is orthogonal and

therefore, independent of the other PC. The output

of the method results in two matrices known as

‘score’ and ‘loadings’. Data were visualised by

plotting the Principal Component (PC) scores,

where each point on the score plot represents an

individual sample. The loading plots were used to

detect spectral areas responsible for separation in the

data, where each point represents one spectral

region. The resulted score values from PCA were

subjected to one way analysis of variance (ANOVA)

to test the levels of statistically significant differences

between treatment and controls. Pattern recognition

of NMR data was performed using in house

developed algorithm on Matlab 7.0 (MathWorks,

Natick, MA, USA). PCA plot generation and

statistical analysis were performed using Sigma plot

11.0 (Systat, San Jose, CA, USA)

Results

1H NMR analysis of serum samples from irradiated mice

A number of perturbations in endogenous metabo-

lites were observed in 1H NMR spectra of serum

samples of irradiated groups collected at different

time points. Figure 1 illustrates 400 MHz 1H NMR

spectra of serum samples from control and irradiated

mice at day 5 post irradiation. The metabolites of

serum affected by radiation exposure are tabulated in

Table I. By visual inspection, at day 1 post

irradiation, few metabolites were altered in group

III compared to controls but no changes had been

observed in Group I and II in respect to controls.

Prominent changes in endogenous serum metabo-

lites, branched chain amino acid (BCA), alanine,

lactate, betaine, choline and phosphoethanolamine,

were observed at day 5 post irradiation in all

irradiated groups compared to controls. Although

the lipid peaks of very low density lipoproteins

(VLDL) and low density lipoprotein (LDL) were

suppressed considerably by CPMG sequence, how-

ever, the peak intensities of VLDL and LDL for

irradiated groups were stronger than those for

control group.

Pattern recognition analysis of serum samples

PCA of NMR spectra was used initially to determine

whether irradiated groups and controls could be

distinguished post irradiation. The PC score plot

from analysis of 1H NMR spectra of serum sampled

at various time points (1, 3 and 5 days post

irradiation) for irradiated mice groups are shown in

Figure 2a–c.

At day 1, 1H NMR spectra and pattern recognition

analysis indicated no obvious change in the serum

samples from mice irradiated with 3 and 5 Gy of

radiation dose. Only mice exposed to 8 Gy could be

separated from controls at day 1 post irradiation.

Significant difference (ANOVA, p5 0.05) on PC1

and PC2 scores were observed between group I, III

and control on day 3 post irradiation. Therefore, at

day 3 time point, mice group irradiated with 3 and 8

Gy dose of radiation could be separated from control

in PC score plot. Remarkable differences (ANOVA,

p5 0.05) between the control and all three irradiated

groups were achieved along PC1 and PC2 axis at day

5 post irradiation that resulted in clear separation of

group I, II and III from control. To compare the

regions of the spectra contributing to the classifica-

tion, PC loading versus chemical shift was plotted in

all groups and control (Figure 3a–c). The PC loading

plots also showed the consistent characterisation

with visual inspection of serum 1H NMR spectra.

For example, the loading plots for irradiated mice

and control group suggested that the separation was

attributed to the NMR signals at d 0.86, 0.9, 0.94,

1.34, 1.48, 2.06, 3.22, 3.54 ppm corresponding to

LDL, VLDL, BCA, lactate, alanine, unsaturated

lipid, choline and glucose which confirmed the visual

comparison of the NMR spectra from irradiated and

control mice (Figure 3).

Discussion

A number of biological compounds become altered

after radiation exposure. These changes occur as a

result of the radiation damage, or in response to

mobilisation for repair, regeneration and cell pro-

liferation (Walden and Farzaneh 1990). Radiation-

induced changes could be reflected as biochemical

perturbations in biofluids. Earlier studies on bio-

chemical and physiological assays have yielded

some important insights into the complex response

Radiation-induced changes in serum by NMR spectroscopy 3

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y In

stitu

te o

f N

ucle

ar M

edic

ine

& A

llied

Sci

enes

on

12/0

9/10

For

pers

onal

use

onl

y.

to ionising radiation, (Altman 1971, Roy et al.

2006). Recently, NMR-based metabonomic meth-

ods have been utilised as a rapid analytical

technique for the study of biochemical variation

in biological fluid (Lindon et al. 2004, Ala-Korpela

2008). In the present study, results of 1H NMR

spectroscopy of serum provide strong evidence of

biochemical variation due to moderate radiation

exposure.

Radiation-induced energy metabolism disturbance and

oxidative stress

In the present study, a number of metabolites

involved in energy generation were perturbed as a

result of gamma irradiation. After radiation expo-

sure, decrease in serum sugar and an increase in

lactate was evident in all three irradiated groups. The

observed decrease in glucose and the concomitant

increase in lactate indicates enhancement in anaero-

bic metabolism. These changes resulted from in-

creased energy demand and were suggestive of

changes in carbohydrate and energy metabolism

after radiation exposure (Lerman et al. 1962).

Increased alanine in serum also suggests a metabolic

switch towards energy conservation and gluconeo-

genesis, which is activated in radiation diseases

(Borovikova et al. 1985). Our observations are in

concurrence with a recent GC-MS based study that

has reported down regulation of energy metabolites,

citrate, 2 oxo glutarate in urine samples in rats post

irradiation (Lanz et al. 2009).

Figure 1. Representative NMR spectra of serum from 3, 5 and 8 Gy irradiated group and control at day 5 post irradiation.

Table I. Summary of proton NMR spectroscopy based-radiation-induced variations in serum metabolites in irradiated mice as compared to

controls.

Metabolites

Chemical shift with

multiplicity (bracket)

3 Gy (Group I) 5 Gy (Group II) 8 Gy (Group III)

Day 1 Day 3 Day 5 Day 1 Day 3 Day 5 Day 1 Day 3 Day 5

Lipids (LDLs/VLDLs) 0.86, 1.30 – " "" – – "" " " ""Branched chain amino acid 0.94 (m) – " " – " "" " " ""Lactate 1.33 (d) – " " – " "" – " ""Alanine 1.48 (d) – " " – " "" – " ""Acetate 1.92 (s) – – – – – – – – –

Glutamine/glutamate 2.46 (m) – – " – – " – – "Choline 3.21 (s) – " " – – " – " ""Phosphoethanol

amine

3.23 (t) – " " – – " – – "

Betaine 3.25 (s) – " " – – "" – – ""Glucose 3.54 (m) – # ## – # ## # # ##

" indicates relative increase in signal; # relative decrease in signal; – no change. Keys: s: singlet; d: doublet; t: triplet; m: multiplet.

4 A. R. Khan et al.

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y In

stitu

te o

f N

ucle

ar M

edic

ine

& A

llied

Sci

enes

on

12/0

9/10

For

pers

onal

use

onl

y.

Ionising radiation generates reactive oxygen spe-

cies (R/OS) as a result of water radiolysis that induce

oxidative stress (Rousselot et al. 1997). These ROS

induce oxidative damage to vital cellular molecules

and structures including lipids, proteins and mem-

branes (Cadet et al. 2004). The increase in amount

of LDL, VLDL, lactate and choline in serum from

irradiated mice was commonly observed in 1H NMR

spectral and pattern recognition analysis. The rise in

the level of serum lipids and lipoproteins on day 5 in

all radiation dose denoted disturbance in lipid

metabolism. Lipoproteins (LDL, VLDL) are impor-

tant carrier proteins of tri acyl glycerol, cholesterol

and cholesteryl esters. Ionising radiation may affect

plasma lipoproteins by several possible mechanisms.

Radiation exposure could modify the interaction of

lipoproteins with their corresponding cell membrane

receptors (Feurgard et al. 1999). This reduced

activity for LDL clearance by the liver may con-

tribute to the marked accumulation of LDL in sera of

irradiated animals. Some earlier studies have shown

increased triacylglycerol levels in plasma due to

inhibition of lipoprotein lipase in small animals

irradiated with g rays (Dousset et al. 1984, Sedlakova

et al. 1986). Many earlier studies have shown

initiation of lipid peroxidation followed by radiation

exposure (Leyko and Bartosz 1986, Pathak et al.

2007). Radiation-induced lipid peroxidation of cell

membrane modifies membrane fluidity (Berroud

et al. 1996) and the activities of some membrane

enzymes (Yukawa et al. 1983). Membranal damage

results in release of phospholipids and choline

compounds as choline is the major head of phos-

pholipids. In our study, an increase in choline and

phosphocholine levels reflects an up-regulation of

synthesis of structural phospholipids to cope with the

demand of membrane resynthesis. Recently, a LC-

MS-based study has reported increased membranous

structure phospholipids and suggested phospholipid

profiles as a promising way to assess the radiation

exposure (Wang et al. 2009). Our results illustrate

changes in membrane metabolites, choline, and

lipoproteins due to radiation-induced oxidative

stress. Similarly, Lanz et al. (2009) have observed

GC-MS-based distinct radiation metabolomic phe-

notype derived from oxidative stress.

Figure 2. Score plots (PC1 vs. PC2) from PCA analysis of proton NMR spectra of serum from control (�), 3 (~), 5 (&) and 8 (.) Gy

irradiated groups at day 1 (a), 3 (b) and 5 (c) post irradiation. Ellipse indicates schematic representation to differentiate the cluster of groups.

Radiation-induced changes in serum by NMR spectroscopy 5

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y In

stitu

te o

f N

ucle

ar M

edic

ine

& A

llied

Sci

enes

on

12/0

9/10

For

pers

onal

use

onl

y.

Radiation-induced amino acid metabolism perturbations

The present study shows increased levels of BCA in

serum of irradiated mice as compared to controls.

Holecek et al. (2002a) have also reported increased

amino acid metabolism at 48 h time point post

irradiation. g-radiation induces degeneration of

skeletal muscle that results in the release of amino

acids from irradiated muscle into blood stream

(Schwenen et al. 1989). As a result of activated

protein breakdown induced by irradiation, BCA are

released from body proteins and extensively catabo-

lised. The amino group released during transamina-

tion of BCA is used primarily for synthesis of alanine

and glutamine (Chang and Goldberg 1978). Alanine

is an important precursor for gluconeogenesis in liver

during increased energy demand. To maintain the

levels of alanine for gluconeogenesis, BCA levels

were released in blood from liver for their catabolism

in skeletal muscle. The observed increase in amino

acid (BCA, alanine) on day 3 and 5 following

irradiation in all three groups reflected the conse-

quences of radiation-induced protein breakdown.

Glutamine acts as a nitrogen shuttle among organs,

an important fuel for rapidly dividing cells such as

erythrocytes and cells of the immune system and a

precursor for the synthesis of nucleotides (Holecek

2002b). However, glutamine was found to be

increased only on day five post irradiation. It shows

that glutamine resynthesis took little more time for

their release in blood stream.

Pattern recognition analysis provides a novel and

powerful method to analyse the data obtained from

NMR spectra after giving respective treatment

(Serkova and Niemann 2006). For classification of

data, PCA has been widely chosen as a preliminary

unsupervised pattern recognition tool to reduce data

dimensions originating from 1H NMR spectra. In

our study, PCA analysis showed both dose and time

dependent changes in the endogenous serum meta-

bolites. Mice treated with higher dose (8 Gy) showed

symptoms at day one post irradiation and could

easily be separated from controls through PCA

analysis. Group I started showing metabolic changes

only at day 3 post irradiation, whereas, group II

could be separated only at day 5 which could be dose

dependent. Similarly, changes observed in irradiated

animals were time dependent. Maximum changes in

all three groups were observed at day 5 post

irradiation, where all three irradiated groups could

be separated from controls at score plot.

Conclusions

From the above findings, it is demonstrated that 1H

NMR spectra of serum samples show metabolic

perturbations in mice after irradiation. In this study,

it is implied that 3, 5 and 8 Gy radiation doses induce

changes in body lipid, protein and membrane

metabolism. Based on these metabolic alterations,

irradiated group could be separated from controls by

NMR-PR analysis. Although the biochemical assays

could not be performed on our samples which could

have correlated better with our results, the study has

identified many earlier known radiation-induced

biochemical changes all together in one spectrum.

This is a basic and first study as per our knowledge to

explore the application of NMR spectroscopy for

studying radiation induced metabolic perturbations

Figure 3. Loading plots from analysis of proton NMR spectra of

serum samples from controls (.) and irradiated groups represent-

ing chemical shift positions corresponding to metabolites altered in

3 (D), 5 (&), and 8 (�) Gy radiation dose-irradiated mice at day 1

(a), 3 (b), and 5 (c) post irradiation.

6 A. R. Khan et al.

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y In

stitu

te o

f N

ucle

ar M

edic

ine

& A

llied

Sci

enes

on

12/0

9/10

For

pers

onal

use

onl

y.

in serum. NMR based metabonomics could be used

for development of metabolic markers and further be

extended for multiparametric approach for mass

screening and evaluation of radiation damage.

Declaration of interest: The authors report no

conflicts of interest. The authors alone are respon-

sible for the content and writing of the paper.

References

Ala-Korpela M. 2008. Critical evaluation of 1H NMR metabo-

nomics of serum as a methodology for disease risk assessment

and diagnostics. Clinical Chemistry and Laboratory Medicine

46:27–42.

Altman KI. 1971. Criteria for the evaluation and selection of

radiation-induced metabolic changes as biochemical indicators

of radiation damage. In: Altman KI, editor. Biochemical

indicators of radiation injury in Man. Vienna: International

Atomic Energy Agency. pp 1–9.

Berroud A, Le Roy A, Voisin P. 1996. Membrane oxidative damage

induced by ionizing radiation detected by fluorescence polariza-

tion. Radiation and Environmental Biophysics 33:455–471.

Bertho JM, Roy L, Souidi M, Benderitter M, Gueguen Y,

Lataillade JJ, Prat M, Fagot T, De Revel T, Gourmelon P.

2008. New biological indicators to evaluate and monitor

radiation- induced damage: An accidental case report. Radia-

tion Research 169:543–550.

Borovikova GV, Mashkova NI, Dokshina GA. 1985. Effect of

tryptophan on gluconeogenesis in the perfused liver of

irradiated rats. Radiobiologia 25:636–639.

Cadet J, Bellson S, Douki T, Frelon S, Gasparutto D, Muller E,

Pouget JP, Ravanat JL, Romieu A, Sauvaigo S. 2004. Radiation

induced DNA damage: Formation measurement and bio-

chemical features. Journal of Environmental Pathology, Tox-

icology and Oncology 23:33–43.

Chang TW, Goldberg AL. 1978. The metabolic fates of amino

acid and the formation of glutamine in skeletal muscle. Journal

of Biological Chemistry 253:3685–3695.

Coleman N, Blakely WF, Fike TJ, MacVitte TJ, Metting NF,

Mitchell JB, Moulder JE, Preston RJ, Ricks RC, Seed TM,

Stone HB, Tofilon PJ, Wong RSL. 2003. Molecular and

cellular biology of moderate-dose (1–10 Gy) radiation and

potential mechanisms of radiation protection: Report of a

workshop at Bethesda, Maryland, December 17–18, 2001.

Radiation Research 159:812–834.

Dousset N, Feliste R, Carton M. and Douste-Blazy L. 1984.

Irradiation induced free cholesterol accumulation in very low

density lipoprotein. Role of lipoprotein lipase deficiency.

Biochimica et Biophysica Acta 794:444–453.

Feurgard C, Boehler N, Ferezou J, Serougne C, Aigueperse J,

Gourmelon P, Lutton C, Mathe D. 1999. Ionizing radiation alters

hepatic cholesterol metabolism and plasma lipoproteins in Syrian

hamster. International Journal of Radiation Biology 75:757–766.

Holecek M, Skopec F, Sprongl L, Mraz J, Skalska H, Pecka M.

2002a. Effect of alanyl-glutamine on leucine and protein

metabolism in irradiated rats. Amino Acids 22:95–108.

Holecek M. 2002b. Relation between glutamine, branched-chain

amino acids, and protein metabolism. Nutrition 18:130–133.

Holmes E, Nicholls AW, Lindon JC, Connor SC, Connelly JC,

Haselden JN, Damment SJ, Spraul M, Neidig P, Nicholson JK.

2000. Chemometric models for toxicity classification based on

NMR spectra of biofluids. Chemical Research Toxicology

13:471–478.

Lanz C, Patterson AD, Slavı́k J, Krausz KW, Ledermann M,

Gonzalez FJ, Idle JR. 2009. Radiation metabolomics 3.

Biomarker discovery in the urine of gamma-irradiated rats

using a simplified metabolomics protocol of gas chromato-

graphy-mass spectrometry combined with random forests

machine learning algorithm. Radiation Research 172:198–

212.

Lerman S, Devi A, Banerjee R. 1962. The effect of ionizing

radiation on protein and carbohydrate metabolism in the rat

lens. Investigative Ophthalmology 1:95–100.

Leyko W, Bartosz G. 1986. Membrane effect of ionizing radiation

and hyperthermia. International Journal of Radiation Biology

49:743–770.

Lindon JC, Holmes E, Bollard ME, Stanley EG, Nicholson JK.

2004. Metabonomic technologies and their applications in

physiological monitoring, drug safety assessment and disease

diagnosis. Biomarkers 9:1–31.

Lindon JC, Nicholson JK, Everett JR. 1999. NMR spectroscopy of

biofluids. Annual Reports on NMR Spectroscopy 38:1–88.

Nicholson JK, Wilson ID. 1989. High resolution proton magnetic

resonance spectroscopy of biological fluids. Progress in

Nuclear Magnetic Resonance Spectroscopy 21:449–501.

Pathak CM, Kumar P, Kumar S, Khanduja KL, Sharma SC.

2007. Whole body exposure to low dose gamma radiation

promotes kidney antioxidant status in Balb/c Mice. Journal of

Radiation Research 48:113–120.

Patterson AD, Li H, Eichler GS, Krausz KW, Weinstein JN,

Forance AJ Jr, Gonzalez FJ, Idle JR. 2008. UPLC-ESI-

TOFMS-based metabolomics and gene expression dynamics

inspector self organizing metabolomic maps as tools for

understanding the cellular response to ionizing radiation.

Analytical Chemistry 80(3):665–674.

Rousselot DB, Khalil A, Delattre J, Jore D, Albert GM. 1997.

Oxidation of human high-density lipoprotein by .OH and

.OH/O.2- free radicals. Radiation Research 147:721–728.

Roy L, Gregoire VE, Durand V, Buard M, Delbos N, Paillole I,

Sorokine-Durm P, Gourmelon, Voisin P. 2006. Study of the

tools available in biological dosimetry to estimate the dose in

case of accidental complex overexposure to ionizing radiation:

The Lilo accident. International Journal of Radiation Biology

82:39–48.

Schwenen M, Altman KI, Schroder W. 1989. Radiation induced

increase in the release of amino acid by isolated perfused

skeletal muscle. International Journal of Radiation Biology

55:257–269.

Sedlakova A, Borikova K, Ahlers I. 1986. Change in lipoprotein

lipase activity in the adipose tissue and heart of non lethally X-

irradiated rats. Physiologia Bohemoslovaca 35:400–405.

Serkova NJ, Niemann CU. 2006. Pattern recognition and

biomarker validation using quantitative 1H-NMR-based

metabolomics. Expert Review of Molecular Diagnostics

6:717–731.

Tyburski JB, Patterson AD, Krausz KW, Slavı́k J, Fornace AJ Jr,

Gonzalez FJ, Idle JR. 2008. Radiation metabolomics. 1. Identi-

fication of minimally invasive urine biomarkers for gamma-

radiation exposure in mice. Radiation Research 170:1–14.

Walden TL, Farzaneh NK. 1990. Biochemistry of ionizing

radiation. New York: Raven Press.

Wang C, Yang J, Nie J. 2009. Plasma phospholipid metabolic

profiling and biomarkers of rat following radiation exposure

based on liquid chromatography-mass spectrometry technique.

Biomedical Chromatography 23:1079–1085.

Yukawa O, Nagatsuka S, Nakasawa T. 1983. Reconstitution

studies on the involvement of radiation induced lipid per

oxidation in damage to membranes enzyme. International

Journal of Radiation Biology 43:391–398.

Radiation-induced changes in serum by NMR spectroscopy 7

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y In

stitu

te o

f N

ucle

ar M

edic

ine

& A

llied

Sci

enes

on

12/0

9/10

For

pers

onal

use

onl

y.

ORIGINAL ARTICLE

NMR spectroscopy based metabolic profiling of urine and serumfor investigation of physiological perturbations during radiationsickness

Ahmad Raza Khan • Poonam Rana • Ritu Tyagi •

Indracanti Prem Kumar • M. Memita Devi •

Salim Javed • Rajendra P. Tripathi • Subash Khushu

Received: 2 October 2010 / Accepted: 7 January 2011

� Springer Science+Business Media, LLC 2011

Abstract Radiation accidents are rare events that induce

radiation syndrome, a complex pathology which is difficult

to treat. In medical management of radiation victims, life

threatening damage to different physiological systems

should be taken into consideration. The present study was

proposed to identify metabolic and physiological perturba-

tions in biofluids of mice during different phases of radiation

sickness using 1H nuclear magnetic resonance (1H NMR)

spectroscopy and pattern recognition (PR) technique. The1H NMR spectra of the biofluids collected from mice irra-

diated with 5 Gray (Gy) at different time points during

radiation sickness were analysed visually and by principal

components analysis. Urine and serum spectral profile

clearly showed altered metabolic profiles during different

phases of radiation sickness. Increased concentration of

urine metabolites viz. citrate, a ketoglutarate, succinate,

hippurate, and trimethylamine during prodromal and clini-

cal manifestation phase of radiation sickness shows altered

gut microflora and energy metabolism. On the other hand,

serum nuclear magnetic resonance (NMR) spectra reflected

changes associated with lipid, energy and membrane

metabolism during radiation sickness. The metabonomic

time trajectory based on PR analysis of 1H NMR spectra of

urine illustrates clear separation of irradiated mice group at

different time points from pre dose. The difference in NMR

spectral profiles depicts the pathophysiological changes and

metabolic disturbances observed during different phases of

radiation sickness, that in turn, demonstrate involvement of

multiple organ dysfunction. This could further be useful in

development of multiparametric approach for better evalu-

ation of radiation damage as well as for medical manage-

ment during radiation sickness.

Keywords Radiation sickness � 1H NMR spectroscopy �Serum � Urine � Metabonomics

1 Introduction

In the present scenario, increasing burden of nuclear arsenal

and scramble of nuclear power leads to a big threat of radi-

ation exposure to the population at large. In all potential

radiation disasters, we are likely to encounter whole body or

partial body radiation exposure. Radiation sickness occurs

after whole body or significant partial body irradiation of

greater than 1 Gray (Gy) radiation dose (Waselenko et al.

2004). It is a sequence of phased symptoms beginning with

prodromal phase of non specific clinical response spanning

from few hours to a few days, followed by symptom free

latent phase lasting a few days or weeks depending upon the

radiation exposure dose. Clinical symptoms reappear during

illness manifestation phase which may last for weeks and

survivability of a radiation exposed person depends on the

recovery in this phase. A dose dependent latent period occurs

between the end of the prodromal phase and the onset of later

haemopoietic or gastrointestinal complications (AFFRI

2003). Therefore, regular monitoring in illness phase is

essential for clinical management. During management of

A. R. Khan � P. Rana � R. Tyagi � M. M. Devi �R. P. Tripathi � S. Khushu (&)

NMR Research Centre, Institute of Nuclear Medicine and Allied

Sciences (INMAS), DRDO, Lucknow Road, Timarpur, Delhi,

India

e-mail: skhushu@yahoo.com

I. P. Kumar

Division of Radiation Biosciences, Institute of Nuclear Medicine

and Allied Sciences (INMAS), Delhi, India

S. Javed

Department of Biochemistry, Jamia Hamdard, New Delhi, India

123

Metabolomics

DOI 10.1007/s11306-011-0277-4

victims of radiological accidents, besides evaluating global

radiation dose received by biological dosimetry, radiation

induced damage to different physiological systems should

also be taken into consideration. In the last few years, new

concept of radiation pathophysiology has been proposed that

involves radiation induced multi-organ involvement (RI-

MOI) and radiation induced multi-organ failure (RIMOF).

Both RIMOI and RIMOF contribute to the clinical outcome

and prognosis of radiation accident victim (Fliedner et al.

2005; GenYao and Changlin 2005). Few bio indicators for

radiation induced damage to the physiological system

(Becciolini et al. 1984; Prat et al. 2006; Lutgens et al. 2004)

are available in literature. Thus there is a crucial need to

obtain information from various metabolic, cellular and

molecular levels of radiation damage during all phases of

radiation sickness. In this scenario, high-throughput non-

invasive studies are required to develop multiparametric

approach with strong prognostic capabilities which could

provide adequate information for subsequent medical man-

agement and treatment decisions (Reeves et al. 1996).1H nuclear magnetic resonance (1H NMR) spectroscopy

is rapidly evolving as a tool in system biology for moni-

toring metabolic fluxes in toxicological studies in small

animals. (Liao et al. 2007; Wei et al. 2008; Wang et al.

2006; Waters et al. 2006). It is well established that nuclear

magnetic resonance (NMR) technique, coupled with pat-

tern recognition (PR) methods, can provide valuable

information on biochemical perturbations due to both

exogenous and endogenous factors by analysis of biofluids

(Waters et al. 2006; Holmes et al. 2008). Serum and urine

samples provide a well established surrogate medium for

the study of metabolic physiological functions of not only

liver and kidney, but also for other vital organs. Earlier

studies in the field of radiation metabolomics have shown

time and dose dependent changes due to acute radiation

stress in animal model and cell system (Tyburski et al.

2008; Lanz et al. 2009; Patterson et al. 2009; Wang et al.

2009; Khan et al. 2010). However, metabolic changes

during different phases of radiation sickness are yet to be

elucidated. The present study has been designed to assess

physiological perturbations in urine and serum during dif-

ferent phases of radiation sickness using 1H NMR spec-

troscopy based metabonomics.

2 Materials and methods

2.1 Animal handling, radiation exposure and sample

collection

Twenty strain ‘A’ male mice of 10 weeks of age were

taken and acclimatized for 48 h in polypropylene cages

prior to group allocation and treatment. They were then

housed individually in metabolic cages in a well-ventilated

room with controlled conditions. During the study, room

temperature was maintained within 19–23�C, relative

humidity within 45–65%, and fluorescent lighting was

provided (6 a.m. to 6 p.m.) for 12 h light/12 h dark cycle.

Mice kept in pi cage were irradiated with 5 Gy of radiation

through gamma irradiation facility at our institute

employing 60Co [Gamma Cell-220, Atomic Energy of

Canada Limited (AECL), Ottawa, Ontario, Canada] with

source operating at 0.20 Gy/min. A dose of 7 Gy in mice is

generally considered to be equivalent to 4 Gy on humans

(Hall and Giaccia 2006). Thus the dose we used was

roughly equivalent to 2.5 Gy on humans and this dose is

considered to be associated with haemopoietic syndrome

with some gastrointestinal symptoms. A summary of

symptoms appearing at different phases of acute radiation

syndrome are given in Table 1.

Urine samples from mice were collected in metabolic

cages. To reduce contamination, mice were placed in clean

cages and 0.1% sodium azide was added in urine collection

tube. To look for metabolic changes during different phases

of radiation sickness, several time points were selected for

sample collection viz. 6 h and day 5 post irradiation for

prodromal phase, day 10 and 15 as latent phase and day 20

and 25 as clinical manifestation phase. For blood collec-

tion, only one time point from each phase of radiation

sickness was selected. Animals (n = 5 at each time point)

were sacrificed and blood was collected through heart

puncture at pre dose, day 5, 10 and 25 post irradiation.

Serum was separated by centrifugation at 2,6659g for

10 min. All samples were stored at -80�C until 1H NMR

spectroscopic analysis was carried out.

2.2 Sample preparation and NMR spectroscopy

All chemicals, NMR solvents, trimethylsilyl-2,2,3,3-tetra-

deuteropropionic acid (TSP) and deuterium oxide (D2O)

used in the study were obtained from Sigma-Aldrich (St

Louis, MO, USA). Animal handling and experimental

protocols adopted in this study were approved by the

Animal Ethical Committee of Institute.

2.2.1 1H NMR spectroscopy measurement of urine

Deuterated phosphate buffer solution (400 ll; 0.2 M

Na2HPO4/0.2 M NaH2PO4, pH 7.4 containing 1 mM TSP

prepared in D2O) was mixed with 200 ll of urine to min-

imize variations in pH of the urine samples. TSP acted as

chemical shift reference (d0.0 ppm) and D2O provided a

lock signal. Samples were then centrifuged at 4,7229g

for 5 min to remove particulate matter and then transferred

to 5 mm NMR tube. NMR spectral data were acquired on

a Bruker-Av400 spectrometer (Bruker, Rheinstetten,

A. R. Khan et al.

123

Karlsruhe Germany) with Broad Band Observe (BBO)

probe operating at a frequency of 400.13 Hz and a tem-

perature of 298 K. Water signal was suppressed using

water presaturation pulse; nuclear overhauser enhanced

spectroscopy (NOESYPR1D) pulse sequence (RD-900-t-

900-tm-900-acq) with relaxation delay (RD) of 2.0 s and an

acquisition time (acq) of 2.5 s was used. Typically, 64 free

induction decay (FID) were collected into 32 K data points

over a spectral width of 6410 Hz.

2.2.2 1H NMR spectroscopy measurement of serum

Serum sample of 200 ll was mixed with 400 ll of D2O

and transferred to 5 mm NMR tubes containing 1 mM TSP

as reference in closed capillary (Wilmad, SP Industries, NJ,

USA). All NMR spectra were acquired at 400.13 MHz on

Bruker-AV400 spectrometer (Bruker, Rheinstetten, Kar-

lsruhe, Germany) at 298 K. Water suppressed Carr–Pur-

cell–Meiboom–Gill (CPMG) spin echo pulse sequence

(RD-900-(s-1800-s)n-acq) with a total spin echo (2ns) of

200 ms was used to attenuate broad signals from proteins

and lipoproteins. Sixty-four FID were collected into 32 K

data points over a spectral width of 8223.68 Hz, with a RD

of 2 s and acq of 3.98 s.

Each FID was weighted by an exponential function with

a 0.3-Hz line broadening factor prior to Fourier transfor-

mation and peaks were assigned on the basis of chemical

shifts according to previously reported literature (Lindon

et al. 1999).

2.3 Data reduction and PR analysis of 1H NMR spectra

All the spectra were phase and baseline corrected and

calibrated (TSP at 0.0 ppm, 1 mM) using TopspinTM

(Bruker, Karlsruhe, Germany). Each 1H NMR spectrum

from urine and serum over the range (d0.2–10.0 ppm) was

reduced to 245 regions of equal width (0.04 ppm) and

signal intensity in each region was integrated using AMIX

software (Bruker, Karlsruhe, Germany). The region

between d4.5 and 5.0 ppm was removed prior to any sta-

tistical analysis in order to elude any spurious effects of

variability in the suppression of water resonance. For urine

spectra, the region containing urea (d5.0–6.0) was also

excluded to eliminate any cross-relaxation effects of urea.

Following removal of these regions, data was normalized

in AMIX by dividing each integrated segment by the total

area of the spectrum to reduce any significant concentration

difference. Output data in ASCII data format was imported

to Microsoft Excel (Microsoft Office 2003), mean centered

and then exported to Matlab 7.1 (MathWorks, MA, USA)

for PR analysis.

Principal component analysis (PCA) was performed for

pre dose and all time points post irradiation urine and

serum samples. PCA plots of first two components

allowed visualisation of the data and to establish whether

there were any time dependent changes in the metabolic

profile of urine or serum post irradiation.Variation

between samples can be detected in the score plots,

whereas spectral region responsible for the differences

can be viewed in the corresponding loading plots. These

spectral regions were examined and potential metabolites

were identified and integrated. Additionally, to calculate

relative changes in identified metabolites during different

phases of radiation sickness, relative concentration of

each identified metabolite in relation to the total spectral

integral area, subsequent to removal of water resonance,

was determined and one way analysis of variance

(ANOVA) and multi comparison test was performed in

MATLAB. The score values from PCA were subjected to

ANOVA to test significant difference between pre and

post irradiation samples.

Mean values of principal component (PC) were calcu-

lated for urine samples at each time point and plots of PC1

versus PC2 for the mean data were constructed. These

maps gave information about the metabolic trajectory

through different phases of radiation sickness.

3 Results and discussion

Radiation incidents are relatively rare events; however, the

consequence of accidental radiation exposure can be

Table 1 Symptoms of acute

radiation syndrome during

exposure of whole body

radiation in the range of 2–6 Gy

of radiation dose

Different phases of radiation

sickness

Symptoms

Prodromal phase Haemopoietic changes: Development of lymphopenia,

granulocytopenia

Gastrointestinal symptoms: Headache, nausea vomiting and diarrhea

Neuromuscular symptoms: Easy fatiguability, fever, headache

Latent phase Pancytopenia and predisposition to infection

Manifest illness phase State of intense immunosuppression

Lymphopenia, opportunistic infections

Haemopoietic and mucosal system affected

Study of radiation sickness by NMR spectroscopy

123

extensive, both physically and physiologically. New

approaches and technologies are available for the study of

health impairments as a function of dose of whole body

irradiation at different organ levels. In this study, we have

used radiation metabonomics based on 1H NMR spectros-

copy to demonstrate radiation induced metabolic or phys-

iological perturbations during different phases of radiation

sickness.

3.1 1H NMR spectroscopic analysis of urine and serum

Typical 1H NMR spectra of urine samples obtained from

irradiated mice at various time points are illustrated in

Figs. 1 and 2. Both types of biofluids contained lactate,

alanine, glutamine, glutamate and alanine. Metabolites like

hippurate, taurine, trimethyl amine (TMA), a-ketoglutarate

(a-KG) and 2-ketoisocaproate that are unique to urine

metabolic profile, were present in urine 1H NMR spectra

amongst others. NMR spectral profile of serum was char-

acterised by predominance of various lipids along with

resonances from creatine, several amino acids and organic

acids.

A number of perturbations in endogenous metabolites

were observed in the 1H NMR spectra of urine samples

collected at various time points post irradiation (Table 2).

Changes were observed mainly in metabolites associated

with energy metabolism (succinate, a-KG, citrate), amino

acids (branched chain amino acids, isoleucine, phenylala-

nine, taurine), gut flora metabolites [hippurate, TMA],

osmolytes (betaine, sarcosine) and creatinine in urine

samples. Figure 3 summarises the relative changes in dif-

ferent metabolites through different phases of radiation

sickness. A dramatic change in metabolic profile occurred

during 6 h post irradiation with significant increase in

almost all metabolites. Continued increase of most of the

metabolites was observed in 1H NMR spectra of urine

samples at day 5. The shift in metabolite changes in irra-

diated animals then turned back to pre dose values at day

10, post irradiation. Thereafter, levels of a-ketoglutarate,

citrate, succinate, trimethylamine and other associated

metabolites of gut flora increased again during day 15, 20

and 25 post irradiation. On the other hand, changes

observed in serum metabolic profile during prodromal and

clinical manifestation illness phase was mainly in lipids,

branched chain amino acid and lactate (Table 3). More-

over, membrane metabolites viz., choline and phospho-

ethanolamine was increased only during day 5 post

irradiation (prodromal phase).

Fig. 1 Comparative 1H NMR urine spectra from irradiated mice at different time points

A. R. Khan et al.

123

3.2 PCA of urine and serum

All urine and serum samples collected at all the time points

of the study were assessed using PCA analysis to investi-

gate time dependent effect of radiation over 25 days of

experimental duration. PCA proved to be a useful and rapid

means of establishing whether the urine spectra obtained

from irradiated mice were different and also identifying at

which time point the maximum biochemical changes have

occurred. A pair wise day to day comparison by PCA can

provide more detailed information on the onset of different

phases of radiation sickness. Score plot of NMR spectra of

urine samples collected during different post irradiation

time point revealed clear separation between pre-dose and

post dose time points (Fig. 4a–f). However, on day 10, post

irradiation, overlapping of score values with pre dose

values suggested that there were no significant changes in

the urine metabolic profile at this time point. PCA score

values calculated at different time points were used to plot

metabolic trajectory, a method of visualising changes in

metabolic profiles with time. PCA analysis based time

trajectory was able to differentiate amongst various phases

of radiation sickness based on physiological perturbations

observed in NMR spectra of urine samples from irradiated

animals. The mean score plots for the first two PCs at each

time point has been shown in Fig. 4g. The trajectory of

irradiated sample moved away from the control or pre dose

position along PC1 axis as early as 6 h and continued till

day 5. Irradiated animals turned back towards pre dose

time point that could be visualised by the close proximity

of day 10 score values to pre dose score. From day 10

onwards, trajectory moved away from pre dose point and

attained maximum shift by the end of the experimental

duration (day 25). In case of serum, aliphatic region

(0.5–4.5 ppm) containing most endogenous metabolite

signals in serum were chosen for PCA analysis. Figure 5

represents score plots on first two PCs of 1H NMR spectra

from pre dose serum samples and samples from irradiated

mice collected on day 5, 10 and 25 post irradiation. Clear

separation of day 5 and day 25 serum samples from pre

dose samples showed radiation induced changes during

prodromal and manifest illness phase respectively. In the

score plot of NMR spectra of serum samples, overlapping

of score values of NMR spectra from serum sample col-

lected on day 10 with pre dose time point sample values

suggest latent phase of radiation sickness. PCA was repe-

ated on group mean data in order to simplify the map and

to establish a biochemical trajectory of effect. The time

Fig. 2 Representative serum 1H NMR spectra from irradiated mice at pre dose, day 5, 10 and 25 post dose

Study of radiation sickness by NMR spectroscopy

123

course of radiation induced changes in serum is shown in

Fig. 5. The metabolic perturbations during radiation sick-

ness phase based on PC loading values and statistical

analysis of spectral regions as described in methods were

characterised by increasing branched amino acids, lactate,

alanine and betaine at day 5 post irradiation. However, no

Table 2 Summary of NMR spectroscopy based radiation induced variations in urine metabolites as compared to pre dose values (identified

based on PCA loading values)

Metabolites Chemical shift with multiplicity (bracket) Prodromal phase Latent phase Manifestation illness phase

6 h Day 5 Day 10 Day 15 Day 20 Day 25

b-Hydroxybutyrate 1.20 (d), 2.31, 2.41 and 4.16 (ABX) : – – – – –

a-Ketoglutarate 2.45 (t), 3.01 (t) :: : – : : :

Succinate 2.41 (s) :: : – : : :

Citrate 2.55 (AB), 2.67 (AB) :: : – : : :

Trimethylamine 2.88 (s) : : : : – :

Creatinine 3.05 (s), 4.06 (s) – : – – – –

Choline 3.21 (s), 3.52 (m) : – – – – –

Betaine 3.27 (s) ; : – – – –

Taurine 3.25 (t), 3.43 (t) – : : : – –

Acetoacetate 2.30 (t), 3.45 (s) : : – – – :

Sarcosine 2.74 (s), 3.60 (s) – : – : – :

Ascorbate 3.74 (d), 3.76 (d), 4.03 (m), 4.52 (d) : : – : : :

Trans-aconitate 3.48 (s), 6.62 (s) : : : : – –

Phenylalanine 3.28 (m), 7.33 (m), 7.38 (m), 7.43 (m) : : – – : :

Hippurate 3.97 (d), 7.55 (t), 7.64 (t), 7.84 (d) – : – : : :

:, indicates relative increase in integral value for the region containing the selected metabolite; –, no change in integral value for the region

containing selected metabolite; ;, indicates relative decrease in integral value for the region containing the selected metabolite. P values for the

changing metabolites were assessed using ANOVA in MATLAB based on the integrals of the selected peaks and were all less than 0.05 levels

s singlet, d doublet, t triplet, m multiplet

Fig. 3 Relative changes in levels of selected metabolites in the

spectra of urine (a–d) and serum (e–f) through different time points of

radiation sickness. Data presented as relative levels with respect to

total spectral area ± standard deviation, with overall differences

between all groups (P \ 0.05) and post hoc tests (P \ 0.05) revealing

a difference from pre dose level

A. R. Khan et al.

123

marked changes could be seen on day 10 post irradiation

(latent phase). During manifest illness phase (day 25) of

study, there was reappearance of increased levels of lipids,

alanine and branched amino acid in NMR spectra of serum.

Our findings corresponded well with different phases of

radiation sickness.

Prodromal phase of radiation sickness is associated with

immediate effect of radiation exposure and lasts for few

hours to few days depending on absorbed radiation dose.

During this acute phase of radiation sickness, most of the

metabolites in serum and urine were found to be increased

after sub lethal dose of whole body radiation. However,

Table 3 Summary of NMR

spectroscopy based radiation

induced variations during

different time points in serum

metabolites as compared to pre

dose values

Metabolites Chemical shift with

multiplicity (bracket)

Day 5 Day 10 Day 25

Lipids (LDLs/VLDLs) 0.86, 1.30 : – ::

Branched chain amino acid 0.94 (m) : – :

Lactate 1.33 (d) : – :

Alanine 1.48 (d) : – :

Acetate 1.92 (s) – – –

Choline 3.21 (s) : – –

Phosphoethanolamine 3.23 (t) : – –

Betaine 3.25 (s) :: – :

Fig. 4 PCA score plots (a–f) based on 1H NMR urine spectra at 6 h,

day 5, 10, 15, 20 and 25 post irradiation, respectively (filled circlepre-dose, open circle post dose). PCA metabolic trajectory plot

(g) summarising the changes in the NMR visible metabolome through

different phases of radiation sickness

Study of radiation sickness by NMR spectroscopy

123

most of the changes observed in urine were Kreb’s cycle

intermediates. These and other energy metabolites repre-

sent bioenergetic status of the body and might have

resulted from increased energy demands and were, there-

fore, suggestive of changes in energy metabolism after

radiation exposure (Lerman et al. 1962). Dose dependent

acute effects of radiation showing altered membrane and

energy metabolism in serum have been reported in our

earlier study (Khan et al. 2010). In this study, highest

increase in the energy intermediates was observed at 6 h

post dose in urine samples. Increased a-keto-glutarate,

succinate and citric acid could also be due to the effect of

radiation on renal tubular physiology. It has been earlier

reported in the literature that whole body radiation causes

depression of tubular secretion and reduces glomerular

filtration in kidney cells (Judas et al. 1997), as a result of

which these metabolites are released in more quantity in

urine. Moreover, mitochondria have long been considered

as direct intracellular target of ionizing radiation which in

turn results in inhibition of citric acid cycle enzyme

(Kergonou et al. 1981). Due to inhibition of these enzymes,

these metabolites might be under utilised in renal tubular

cells and this could have resulted in release of these

metabolites in urine of irradiated mice. Inhibition of aer-

obic mitochondrial enzyme or metabolism is also sup-

ported with the presence of increased lactate in serum

NMR spectra of irradiated mice. Therefore, changes

observed in energy intermediates at 6 h post dose could be

cumulative expression of both disturbed energy metabo-

lism due to increased energy demands as well as altered

renal tubular function.1H NMR spectra of serum from irradiated mice showed

radiation induced oxidative stress related changes in the

form of increased choline compounds. Ionizing radiation

generates reactive oxygen species (ROS) as a result of

water radiolysis that induce oxidative stress (Rousselot

et al. 1997). These ROS induce oxidative damage to vital

cellular molecules and initiate lipid peroxidation (Cadet

et al. 2004). Membranal damage due to lipid peroxidation

results in release of phospholipids and choline compounds

as choline is the major head of phospholipids. In our study,

increase in choline and phosphocholine levels in NMR

spectra of serum at day 5 reflects an up-regulation of

synthesis of structural phospholipids to cope with the

demands of membrane resynthesis. However, this altered

membrane metabolism was restricted to only prodromal

phase, since most of the changes occurring in this phase are

due to radiation induced oxidative stress.

Ionizing radiation is said to cause oxidative damage to

tissues and a few earlier studies have reported radiation

induced liver and kidney injury (Jindal et al. 2006; Sharma

et al. 2006; Rana et al. 2009). In our study, radiation

induced changes in lipid metabolism were also observed in

the form of increased lipoproteins viz. low density lipo-

protein and very low density lipoprotein (LDL and VLDL

respectively) in serum 1H NMR spectral analysis. Signifi-

cant increase in the levels of serum lipid profile and LDL

are demonstrated 5 days post irradiation possibly as a

result of liver injury which is in agreement with previous

studies in Syrian hamster (Feurgard et al. 1999) and mice

(Agrawal et al. 2001). This indicates that ionizing radation

induced oxidative stress might alter hepatic lipid metabo-

lism and serum lipoproteins. Radiation induced liver injury

has also been reflected as increased taurine signals in urine

metabolic profile in our study. Taurine is considered as

urinary biomarker for liver injury (Waterfield et al. 1993;

Beckwith-Hall et al. 1998; Clayton et al. 2003) and its level

was increased during prodromal phase which continued to

increase till day 15. Increased urinary taurine is generally

observed in fatty liver and in our study liver steatosis was

observed in histological studies on liver tissue obtained on

day 5 from mice irradiated with a radiation dose of 5 Gy

(unpublished observations).

Acute radiation sickness involves gastrointestinal

symptoms that include diarrhoea (Table 1). Gastrointesti-

nal symptoms during radiation sickness could be due to

radiation induced dysmotility of intestine, denudation of

villi and loss of epithelial lining that could be associated

with bacterial overgrowth (Macnaughton 2000; Somsoy

et al. 2002). Bacterial overgrowth during radiation infec-

tion might be related with disturbed gut flora metabolism.

Gut microbiota extensively catabolise proteins and aro-

matic amino acids such as hippurate, phenylacetylglycine,

phenylalanine, tyrosine and tryptophan. These species are

present in urine as products of intestinal microflora

Fig. 5 A plot of PC1 versus PC2 based on 1H NMR integral region

for individual serum samples at pre dose (filled circle), day 5 (filledsquare), 10 (open square) and 25 (open circle) post dose obtained

from irradiated mice and mean data of samples (cross) showing time

related trajectory

A. R. Khan et al.

123

metabolism. Disturbed gut flora metabolism is directly

reflected in urine NMR spectral profile as increased TMA,

hippurate, phenylalanine and other aromatic amino acid

signals. TMA is an indicator of methylamine metabolism.

Methylamines are important osmo-regulatory compounds

and are produced via degradation of dietary choline to

TMA and its di and mono amine metabolites by gut flora

(Martin et al. 2008). In our study, increased levels of TMA

might be attributed to altered intestinal bacterial metabo-

lism. The increased levels of urinary phenylalanine and

phenylacetylglycine related compounds and hippurate also

suggest radiation induced effects on the gut flora because

their precursors, benzoic acid and phenylacetic acid are

produced by bacterial metabolism.

Maximum radiation induced changes appeared during

prodromal phase as radiation induced oxidative stress lasts

mainly in this phase. In latent phase, reversal of most of the

changes occurring during prodromal phase to normal could

be due to repair mechanism occurring simultaneously in

the body. However, changes started reappearing from day

15 onwards in manifest illness phase and reappearance of

these changes during clinical manifestation illness phase

could be associated with immunosuppressed state of the

body during lymphocytopenia or other changes in haemo-

poietic and gastrointestinal system during prodromal phase.

Although, these changes in metabolic profile were not as

high as observed in prodromal phase of radiation sickness.

In our study, most of the changes during late phase pertain

mainly to gut flora and energy metabolism suggesting

disturbed gastrointestinal symptoms and opportunistic

infection due to immunosuppresion of the body.

Our NMR based spectral profiling of biofluids has

identified changes in organal and physiological functioning

at different phases of radiation syndromes. In recent years,

new concept of radiation pathophysiology has been pro-

posed that involves multi organ injury, multi organ dys-

function or multi organ failure and marked involvement of

lung, kidney, liver and other organs in addition to acute

radiation syndrome (Fliedner et al. 2005; Waselenko et al.

2004). Although we have not observed clinical parameters

of liver, kidney or other organ dysfunction, our results

suggest patho-physiological changes leading to multi organ

dysfunction syndrome. Consequently, it could be proposed

that an underlying radiation induced multi organ dysfunc-

tion syndrome should be evaluated irrespective of radiation

dose exposure.

4 Conclusions

The application of metabolomics in radiation casualty

management and monitoring is still in its infancy. How-

ever, information provided using metabolomics will

nonetheless aid in medical diagnostic practices during

radiation accident. Results of the present study clearly

showed metabolic changes during different phases of

radiation sickness. These changes could be the conse-

quence of radiation induced damage to physiological sys-

tem leading to multi organ dysfunction. Such studies will

not only help in medical management during radiation

accident but also beneficial for patients undergoing cancer

treatment using ionizing radiation.

Acknowledgments This work was peformed as part of DRDO

sponsored R & D project INM 308. The aurhors are grateful for the

financial support from Defence Research & Development Organisa-

tion (DRDO), Ministry of Defence, India.

References

AFFRI. (2003). Handbook on medical management of radiologicalcasualties (2nd ed.). Bethesda, USA: Military Medical Opera-

tion, Armed Force Radiobiology Research Institute (AFFRI).

Agrawal, A., Chandra, D., & Kale, R. K. (2001). Radiation induced

oxidative stress: II. Studies in liver as a distant organ of tumor

bearing mice. Molecular and Cellular Biochemistry, 224, 9–17.

Becciolini, A., Giannardi, G., Cionini, L., et al. (1984). Plasma

amylase activity as biochemical indicator of radiation injury to

salivary glands. Acta Radiologica Oncology, 23, 9–14.

Beckwith-Hall, B. M., Nicholson, J. K., Nicholls, A. W., et al. (1998).

Nuclear magnetic resonance spectroscopic and principal com-

ponents analysis investigations into biochemical effects of three

model hepatotoxins. Chemical Research in Toxicology, 11,

260–272.

Cadet, J., Bellson, S., Douki, T., et al. (2004). Radiation induced

DNA damage: Formation measurement and biochemical fea-

tures. Journal of Environmental Pathology, Toxicology andOncology, 23, 33–43.

Clayton, T. A., Lindon, J. C., Everett, J. R., et al. (2003). An

hypothesis for a mechanism underlying hepatotoxin induced

hypercreatinuria. Archives of Toxicology, 77, 208–217.

Feurgard, C., Boehler, N., & Ferezou, J. (1999). Ionizing radiation

alters hepatic cholesterol metabolism and plasma lipoproteins in

Syrian hamster. International Journal of Radiation Biology, 75,

757–766.

Fliedner, T. M., Dorr, H. D., & Meineke, V. (2005). Multiorgan

involvement as apathogenetic principle of the radiation syn-

drome: A study involving 110 case histories documented in

search and classified as the bases of haematopoetic indicators of

effect. British Journal of Radiology, 27, 1–8.

GenYao, Y., & Changlin, Y. (2005). Multiorgan involvement and

failure in radiation accident: The Chinese experience of 1963.

British Journal of Radiology, 27, 55–61.

Hall, E. J., & Giaccia, A. J. (2006). Radiobiology for the radiologist.Philadelphia: Lippincott Williams & Wilkins.

Holmes, E., Loo, R. L., Stamler, J., et al. (2008). Human metabolic

phenotype diversity and its association with diet and blood

pressure. Nature, 453, 396–401.

Jindal, A., Soyal, D., Singh, I., et al. (2006). Modification of

radiation-induced damage in mice by Rosemarinus officinalisextract (roe). Pharmacology Online, 2, 63–75.

Judas, L., Bentzen, S. M., & Stewart, F. A. (1997). Progression rate of

radiation damage to the mouse kidney: A quantitative analysis of

experimental data using a simple mathematical model of the

Study of radiation sickness by NMR spectroscopy

123

nephron. International Journal of Radiation Biology, 72,

461–473.

Kergonou, J. F., Braquet, M., & Rocquet, G. (1981). Influence of

whole-body irradiation upon rat liver mitochondrial fractions.

Radiation Research, 88, 377–384.

Khan, A. R., Rana, P., Devi, M., et al. (2010). Nuclear magnetic

resonance spectroscopy-based metabonomic investigation of

biochemical effects in serum of c irradiated mice. InternationalJournal of Radiation Biology (in press).

Lanz, C., Patterson, D. A., Slavı́k, J., et al. (2009). Radiation

metabolomics. 3. Biomarker discovery in the urine of gamma-

irradiated rats using a simplified metabolomics protocol of gas

chromatography-mass spectrometry combined with random

forests machine learning algorithm. Radiation Research, 172,

198–212.

Lerman, S., Devi, A., & Banerjee, R. (1962). The effect of ionizing

radiation on protein and carbohydrate metabolism in the rat lens.

Investigative Ophthalmology, 1, 95–100.

Liao, P., Wei, L., & Zhang, X. (2007). Metabolic profiling of serum

from gadolinium chloride treated rats by 1H NMR spectroscopy.

Analytical Biochemistry, 364, 112–121.

Lindon, J. C., Nicholson, J. K., & Everett, J. R. (1999). NMR

spectroscopy of biofluids. Annual Reports on NMR Spectros-copy, 38, 1–88.

Lutgens, L., Deutz, N., Granzier-Peeters, M., et al. (2004). Plasma

citrulline concentration: A surrogate endpoint for radiation

induced mucosal atrophy of the small bowel. A feasibility study

in 23 patients. International Journal of Radiation Oncology,Biology, Physics, 60, 275–278.

Macnaughton, W. K. (2000). Review article: New insights into the

pathogenesis of radiation-induced intestinal dysfunction. Ali-mentary Pharmacology & Therapeutics, 14, 523–528.

Martin, F.-P. J., Wang, Y., Sprenger, N., et al. (2008). Probiotic

modulation of symbiotic gut microbial–host metabolic interac-

tions in a humanized microbiome mouse model. MolecularSystems Biology, 157, 1–15.

Patterson, A. D., Lanz, C., Gonzalez, F. J., et al. (2009). The role of mass

spectrometry-based metabolomics in medical countermeasures

against radiation. Mass Spectrometry Reviews, 29, 503–521.

Prat, M., Demarquay, C., Frick, J., et al. (2006). Use of FLT ligand to

evaluate residual hematopoiesis after heterogenous irradiation in

mice. Radiation Research, 166, 504–511.

Rana, P., Singh, P., Khan, A. R., et al. (2009). High resolution MR

spectroscopy reveals radiation induced metabolic changes in

mouse liver tissues. In Proceedings of the international societyfor magnetic resonance in medicine, Honolulu, HI

Reeves, G. I., David, G. J., Thomas, M. S., et al. (1996). Triage of

irradiated personnel. In An Armed Force Radiobiology ResearchInstitute workshop, September 25–27, 1996.

Rousselot, D. B., Khalil, A., Delattre, J., et al. (1997). Oxidation of

human high-density lipoprotein by �OH and �OH/O�2- free

radicals. Radiation Research, 147, 721–728.

Sharma, M., Mccarthy, E. T., Sharma, R., et al. (2006). Arachidonic

acid metabolites mediate the radiation-induced increase in

glomerular albumin permeability. Experimental Biology andMedicine, 231, 99–106.

Somsoy, Z., Horvath, G., Telbisz, A., Rez, G., & Palfia, Z. (2002).

Morphological aspect of ionizing radiation response of small

intestine. Micron, 33, 167–178.

Tyburski, J. B., Patterson, D. A., & Krausza, W. K. (2008). Radiation

metabolomics. 1. Identification of minimally invasive urine

biomarkers for gamma-radiation exposure in mice. RadiationResearch, 170, 1–14.

Wang, Q., Jiang, Y., Wu, C., et al. (2006). Study of novel indolin-2-

ketone compound Z24 induced hepatotoxicity by NMR-spec-

troscopy-based metabonomics of rat urine, blood plasma and

liver extract. Toxicology and Applied Pharmacology, 215,

71–82.

Wang, C., Yang, J., & Niea, J. (2009). Plasma phospholipids

metabolic profiling and biomarkers of rats following radiation

exposure based on liquid chromatography–mass spectrometry

technique. Biomedical Chromatography, 23, 1079–1085.

Waselenko, J. K., MacVittie, T. J., Blakely, W. F., et al. (2004).

Medical management of the acute radiation syndrome: Recom-

mendations of the strategic national stockpile radiation working

group. Annals of Internal Medicine, 140, 1037–1051.

Waterfield, C. J., Turton, J. A., Scales, M. D., et al. (1993). The

correlation between urinary and liver taurine levels and between

pre-dose urinary taurine and liver damage. Toxicology, 77, 1–5.

Waters, N. J., Waterfield, C. J., Farrant, R. D., et al. (2006). Integrated

metabonomic analysis of bromobenzene-induced hepatotoxicity:

Novel induction of 5-oxoprolinosis. Journal of ProteomeResearch, 5, 1448–1459.

Wei, L., Liao, P., Wu, H., et al. (2008). Toxicological effects of

cinnabar in rats by NMR-based metabolic profiling of urine and

serum. Toxicology and Applied Pharmacology, 227, 417–429.

A. R. Khan et al.

123