See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/259641327

Effects of the environmental stress on two fish populations revealed by

statistical and spectral analysis

Article in Environmental engineering and management journal · January 2012

DOI: 10.30638/eemj.2012.016

CITATIONS READS

9 318

9 authors, including:

Gabriel Lazar Dorel Ureche

University of Bacau University of Bacau
137 PUBLICATIONS 451 CITATIONS 27 PUBLICATIONS 84 CITATIONS

SEE PROFILE SEE PROFILE

Marius Stamate Camelia Ureche

Young Promotion NGO University of Bacau
46 PUBLICATIONS 335 CITATIONS 23 PUBLICATIONS 41 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Aluminosilicate-based porous materials for adsorption and catalysis View project

Health and environmental risks related to the groundwater and soil contamination with organochlorine pesticides and heavy metals View project

All content following this page was uploaded by Gabriel Lazar on 05 April 2014.

The user has requested enhancement of the downloaded file.

Environmental Engineering and Management Journal January 2012, Vol.11, No. 1, 109-124
http://omicron.ch.tuiasi.ro/EEMJ/

“Gheorghe Asachi” Technical University of Iasi, Romania

EFFECTS OF THE ENVIRONMENTAL STRESS ON TWO

FISH POPULATIONS REVEALED BY STATISTICAL
AND SPECTRAL ANALYSIS

Gabriel Lazar1, Dorel Ureche1, Irina Loredana Ifrim1, Marius Stamate1,

Camelia Ureche1, Valentin Nedeff1, Ileana Denisa Nistor1,
Adriana Luminita Finaru1, Iuliana Mihaela Lazar1,2
1
“Vasile Alecsandri” University of Bacau, 157 Calea Mărăşeşti, 600115 Bacău, Romania
2
“Carol Davila” University of Medicine and Pharmacy, Biophysics and Cell Biotechnology Department,
8 Eroii Sanitari Blvd., 050461 Bucharest, Romania

Abstract

The aquatic species are permanently subjected to various stresses through the ecosystem and the food habit. Elucidating the
impact at the cellular level allows us to understand the ways in which the polluting elements and their derivatives can disrupt the
metabolism of aquatic organisms triggering apoptosis in individual cells or necrosis affecting groups of neighbouring cells. The
aim of this preliminary study is to highlight the influence of the environmental factors on two fish population by statistical and
spectroscopic methods, allowing the observation of the changes in structural components of cell’s membranes. In many studies, an
accurate method for biomolecules quantification is necessary for the purpose of selecting optimum species and environmental
living conditions. Biological samples were collected from two different ecosystems: the Oituz River placed in a region less
affected by urban and industrial pollution, and the Mures River, located in an area with a higher degree of pollution. Different
samples from brain and muscle and theirs etheric extract were analyzed using ATR-FTIR and UV-VIS spectrometry. Detailed
analysis of the different spectral regions after spectral decomposition reveals differences between the two fish populations. The
infrared spectroscopy was assisted by statistical analyses in order to develop a calibration model and to reveal the mainly spectral
region that correlate with the etheric extract.

Key words: ATR-FTIR, chemometric, environment, fish, UV

Received: September, 2011; Revised final: January 2012; Accepted: January, 2012

1. Introduction FTIR spectroscopy due to fine details and

structural information in fish sample analysis has
An increasing number of chemicals are been useful in monitoring biochemical and
discharged into the environment, many of which are biophysical changes at cell membrane level (Laurens
dangerous for living organisms and ecosystems. Fish, and Wolfrum, 2011). The application of the applied
as living bioindicator species, are used extensively as screening method can be optimized for mutant or
biological screens of environmental variations of transgenic lines from one fish species or a multiple
anthropogenic pollutants levels. fish species combined from different places.
Fish plays important roles as indicator of trace It is known that the heavy metals can be linked
element pollution, because they occupy different to proteins and membrane lipids, altering their
trophic levels, and are different in size and age structure and function. The cell membrane is the first
(Palaniappan et al., 2010). target structure responding through a specific receptor


Author to whom all correspondence should be addressed: e-mail: [email protected]; Phone: +40 234542411; Fax: +40 234545753
 Lazar et al./Environmental Engineering and Management Journal 11 (2012), 1, 109-124

protein when the pollutants penetrate the cell through each organ system at the molecular level that is
different transport mechanisms, like endocytosis affected.
(Moisescu et al., 2008). When the production rate of FTIR spectroscopy was applied to monitor the
reactive oxygen and nitrogen species exceeds the rate structural changes in lipids and in amino acids of
of decomposition induced by antioxidants, the proteins (Iwaki et al., 2003), to study the damage and
oxidative stress appears, modifying the metabolism of coordination effects of heavy metals on photosystem
cellular macromolecules. II proteins (Nahar and Tajmir-Riahi, 1996), to analyze
The common methods for monitoring the role of the metallothioneins (MT) (Coyle et al.,
parameters like oxidation of lipids in fish population 2002; Formigari et al., 2007; Vergani et al., 2005), to
are based on chemical analysis. The oxidative study the enzymatic activities in the detoxification
deterioration of lipids in fish oil were monitored using mechanism in response to the presence of different
reflectance NIR and absorbance MIR spectroscopy pollutants in the aquatic environment (Henczová et
accompanied by PLS regression. Hydrolytic al., 2008). Other advantages of FTIR spectroscopy
deterioration of lipids from fish oil, which seriously were to detect abnormal changes in cells and tissues
affects the use of oil and industry storage with minimal preparation under in vivo conditions
requirements, may be successfully monitored by during minimal invasive investigation (Krafft et al.,
spectroscopy and chemometric methods in fishmeal 2004).
industry (Cozzolino et al., 2005). Additionally, it was The infrared spectroscopy has been used by
also possible to achieve reliable estimates of fat Laurens and Wolfrum (2011) to develop the
content using the middle infrared signal intensity multivariate chemometric model for estimating levels
from 1744 cm-1 (Sun, 2009). of spiked neutral and polar fats from micro algae.
Other methods currently available for Benefit of infrared spectroscopic techniques is the
screening lipid content in different sample are based nature straight, fast and non-destructive method of
on fluorescent lipophilic dyes such as Nile red, thin screening for specific characteristics of a large
layer chromatography (TLC), and mass spectroscopy number of samples bypassing the need for long and
(MS). UV-VIS Diffuse Reflectance Spectroscopy laborious wet chemical analyses.
(Arsene et al., 2010), Fourier Transform Infrared The calibration and validation model provides
(FTIR) (Harja et al., 2010) are non-invasive a quick way and high throughput to determinate the
alternative techniques to study the molecular structure content of lipid, offering an alternative to other
and composition modifications concerning a large methods. A multivariate partial least squares (PLS)
variety of samples. A mathematical model to predict like chemometric approach was used to correlate the
the oxidative values of Menhaden fish oil, the main spectral changes during increasing
artificial neural network analysis (ANN) (Gosav et phospholipids’ and triglyceride content in algal
al., 2007), was used by Klaypradit to evaluate biomass collected and to predict the levels of
oxidative quality with ATR technique (Klaypradit et exogenously added lipids of trilaurin as a triglyceride
al., 2010). ATR spectroscopy along with ANN proves and phosphatidylcholine as a phospholipids (Laurens
a rapid technique instead of the conventional method and Wolfrum, 2011).
for prediction of fish lipids oxidation. Levels of antioxidant protection and lipid
Heavy metals (Hlihor and Gavrilescu, 2009; peroxidation were evaluated in different species of
Stefanescu et al., 2011) present in some industrial marine organisms (Almeida et al., 2004). Mussels
wastewaters, despite the essential role of some of the exposed to Cd, Cu, Fe, Pb for different time periods
metals in a number of enzymatic processes are increased lipid peroxidation that could share in low
potentially toxic to freshwater environments levels of reduced glutathione (GSH) and glutathione
(Teodosiu et al., 2009) because of toxic effects on the peroxidase (GPX) activity and depletion of GSH and
receiver, influencing the performance of biological caused GPX activity increased. The correlation
treatment processes (Aksu, 2005). Another FTIR coefficients between the enzyme phospholipid
spectroscopy application was to detect the damaging hydroperoxide glutathione peroxidase PHGPx and
effects of the heavy metals on carp (Cyprinus carpio malonaldehyde levels were performed with statistical
L.), silver carp (Hypopthalmichtys molitrix V.) and software. Negative correlation suggested that enzyme
wels (Silurus glanis L.). To detect the damaging PHGPx was used as a biomarker of the toxicity of
effects of the heavy metals, the investigated fish new potential pollutants related with exposure to the
isoenzymes showed that the Cu2+ was a stronger mussels. Peroxidase lipid is correlated with
inhibitor than Pb2+ (Henczová et al., 2008). antioxidant activity classical systems (Almeida et al.,
The high Cu2+ concentrations appear to be 2004).
toxic to fish, especially if it accumulates beyond the Therefore, correlating macroscopic with
chelating capacity of the cell, generating oxidative microscopic and structural properties is a key element
stress and cellular damage (Gaetke and Chow, 2003; in models development. Chemometrical processing of
Henczová et al., 2008). The continuous exposure of measured data allows the detection of the most
marine organisms to Pb2+ low concentrations may important spectral features, which are crucial for
result in their bioaccumulation and subsequent ranking the fish samples into several groups
transfer to humans through the food chain, owed for according to their specific geographical region.

110
 Effects of the environmental stress on two fish populations revealed by statistical and spectral analysis

In the presented study such chemometric using a ZnSe reflectance cell. Samples were carefully
analysis methods are performed using Unscrambler X homogenized and were pressed against the ZnSe cell
10 and SPSS 17.0 software to obtain information with prior to scanning.
minimum error from a reduced number of fish A total of 32 scans were taken for each
samples whose macroscopic characteristics are spectrum, for three replicate samples. The spectra
known. The uses of chemometric method in were collected in the range of 4000 to 600 cm−1 at 1
combination with FTIR spectroscopy are cm−1 resolution and data were exported using Opus
recommended for the discrimination of two fish 6.0 software. Infrared measurements were performed
populations. The main objective of this article is to on samples taken from muscles and brain, carefully
interpret the IR and UV-VIS absorption spectra with homogenized. Background spectra, which were
deconvoluted absorption peaks highlighting the collected under identical conditions, were subtracted
differences between the fish samples. The second from the brain and muscle sample spectra
objective is to develop a rapid calibration method to automatically. All measurements were made at the
predict the percent of lipid extracts in different fish laboratory temperature 22°C. Spectra were recorded
species. on samples before extraction, after extraction of fatty
acids and on extracted samples.
2. Material and methods The extraction was performed using a
microwave equipment Milestone ETHOS lab station,
2.1. Biological material a high-performance microwave-accelerated system
with advanced process control. This method provides
Biological samples were collected from two the solvent extraction of fatty acid from the samples
different ecological environments: in a closed vessel device using temperature control
- Oituz River, Oituz collection point: GPS microwave heating. Diethyl ether was used as a
Coordinates: N: 46.15063 E: 26.45176 Altitude: 200 solvent. Extraction takes place at 35°C and 800 W for
m; covered area: 527 m2; type of course: linear; 20 minutes.
structure of the substrate: gravel, sand, silt, riparian; UV-Vis spectra were recorded only for
vegetation: shrubs, herbaceous plants; aquatic extracted samples using a Varian Cary 100
vegetation: periphyton; physicochemical Spectrophotometer. The molecular absorption spectra
characteristics: air/water temperature: of etheric extract from muscle and brain of the
29.10°C/25.30°C, pH: 8.55; fishing date: 03.09.2008; investigated fish population were measured in the
fishing time: 17 h 10 '; region of 200-400 nm. The absorbance values at 200
- Mureş River, Mureş Basin, Vidrasău collection wavelengths were used as spectral variables for
point: GPS coordinates: N: 46.28877 E: 24.25147 statistical analysis.
Altitude: 285 m; covered area: 228 m2; type of
course: sinuous; structure of the substrate: gravel, 2.3. Spectral processing of middle infrared and UV-
sand; riparian vegetation: trees, shrubs, herbaceous VIS experimental data
plants, reed; aquatic vegetation: periphyton;
physicochemical characteristics: air/water The FTIR instrument provides software for
temperature: 28.10°C/20.10°C, pH: 8.278; fishing spectral subtraction, smoothing, derivative and
date: 10.09.2009; fishing time: 16 h 50 '. deconvolution analysis. The availability of software
The biological material was collected by plays an important role in analyzing of subtle changes
electroshock narcosis, using an electric fishing in infrared biomolecules (Barth and Haris, 2009). For
machine with double insulation EGF 5000 (EFKO - FTIR models the mathematical pre-treatment is an
Elektrofischfanggerate GmbH, Germany). The improvement over the models using the raw spectra.
collected material was fixed with formalin 4%. In our study the FTIR spectral frequency has
Conservation has been achieved in technical alcohol, been achieved with the Opus Bruker software. The
water and glycerin in a ratio of 1: 2: 0.25 (Nicoară et obtained data were exported to Origin 8.0 software.
al., 2006). The mathematical treatment of spectral data was
In Romania the Squalius cephalus L., 1758 made before making the calibration models for
(chub) species lives almost exclusively in small rivers chemometric analyzed. This pre-treatment can help to
and rarely in ponds. In Europe chub live north of the reduce spectral variation due to instrument or sample
Pyrenees, Alps, Dinaric and Balkan. The Barbus variability (Laurens and Wolfrum, 2011).
petenyi Heckel, 1847 species lives only in rivers and Attention is required because the mathematical
streams from mountains and the upper region of hills. pre-treatment of the raw spectra affects the quality of
Both species live in rocky rivers, fast and cold, and in the prediction models. All spectra were smoothed
some streams where summer heat is strong, but only with Savitzky-Golay method at 9 points of window
in the mountains. and 2 polynomial orders. The areas of the absorption
intensity of peaks which are directly related to the
2.2. FTIR and UV-VIS spectroscopic method concentration of the molecules were calculated using
the base line correction (Venkataramana et al., 2010).
FTIR attenuated total reflectance spectra were The fitting curve for FTIR and UV-VIS
collected with a Tensor 27 Bruker spectrophotometer, spectra data were made with Gaussian peaks type.

111
 Lazar et al./Environmental Engineering and Management Journal 11 (2012), 1, 109-124

UV-VIS spectra were smoothed with Savitzky-Golay measuring parameters and electrical noise. Among
method at 15 points of window and polynomial of these factors, the greatest influence is given by
order 3. R2 value of fitting curve for all samples were changes of concentrations of mixture constituents.
≈ 0.9999 value and Reduced Chi Sq ≈ 8·10-5. The Thus, using the complex and sophisticated
value of R2 and Reduced Chi Sq were represented on mathematical methods, which explanation exceeds
each graph. this paper, based on calibration sample spectra helps
achieving a more limited set of principal components
2.4. Multivariate calibration or PCs factors. Representation of the samples in the
PCs space gives us a set of scores and loadings for
Generally, constituents of a sample have each PCs. So, based on calibration sample spectra,
numerous bands that are partially or completely were achieved a more restricted set of main
overlapping. However, based on Lambert - Beer law components or factors that include 90-99% of
it can be taken into account that mixtures’ absorbance variables values.
is additive. Therefore, for any two components m and
n, the equations bases of the calibration process are 2.4.2. Partial Least Squares (PLS)
Eqs. (1) and (2) (Neamtu and Anoaica, 2006). The Partial Least Squares (PLS) method
performs correlations between spectrum
X i  ki mCm  ki nCn (1) decomposition and concentration of mixture’s
components. The method has two different
X  j  k  j m C m  k  j n Cn (2) algorithms: PLS1 and PLS2. PLS1 calculates one set
of spectral factors and scores for each mixture’s
constituent of interest. PLS2 provides a spectral
where: kλ is a constant which includes the extinction
decomposition for all constituents in the mixture. This
coefficient and the specific optical path way for each
means that in this case, it will be obtained a set of
component m and n, and λi, λj are the wavelengths
spectral factors and one of scores for whole set of
corresponding to absorption bands and Cm, n is the
calibration spectra. The main advantages of the PLS
concentration of m and n constituents that have
methods are: the decomposition and regression takes
numerous partially or totally overlap bands.
place in one step, derived spectral factors are directly
Using data measured at multiple wavelengths
linked to constituents of the mixture.
for a great number of samples with known properties,
The method can be applied to highly complex
there is a possibility to minimize the errors, so the
mixtures, in which it is enough to know only the
measurement results are almost near the real value.
concentration of calibration constituents. The
Specific calculation methods consist of writing the
calibration model can be used to predict contaminated
linear algebra matrix of the system equations
samples with constituents that were not present in the
characterizing the complex sample analyzed in the
calibration process. The advantage of these methods
infrared or in the ultraviolet-visible region.
is the possibility of using the wavelengths in the
Multivariate spectrometry, name also
entire range in the calibration method, thus achieving
chemometric method, is based on the specific spectral
an effect of maintenance on the experimental results,
mathematical dependence of absorption intensity of
enhancing the accuracy of the calibration model.
elements, radicals, functional groups or molecules
from simple compounds or complex mixtures of
2.5. Correlations and regression analysis
compounds and their concentrations. This method
applies particularly when the theory available at that
The chemometric method is a less expensive
time is not sufficient for solving problems that have
approach because it is not necessary to work with the
many unknown variables in relationships.
corresponding standards to confirm the outputs of
The chemometric strategy consists in
qualitative analysis performed by a chosen spectral
following steps:
method. It is not necessary to assign all spectral bands
- data acquisition;
or peaks. In this study, the correlations between the
- generate a mathematical model that is based on
percent of lipid extract and the relative intensity area
statistical methods or neural networks;
of middle infrared spectral regions were investigated.
- interpretation of model parameters achieved;
For this purpose the Bivariate Correlations procedure
- applying the model to a new situation.
was computed and Pearson's correlation coefficient
was calculated with SPSS 17.0 software.
2.4.1. Principal Component Analysis PCA
These variables are negatively or positively
The method of Principal Components Analysis
correlated and the correlation is significant at the 0.05
(PCA) underlies chemometric analysis of several
level. General aim of multiple regressions is to
quantitative and qualitative spectrophotometric
observe the relationship between several dependent
methods.
and independent variable.
Note that a sample spectrum is obtained as a
In this case the dependent variables are:
result of various factors changes such as constituents
phosphatidylcholine, deformation of carbohydrates,
of mixture, interactions between constituents, change
sphingolipids&oligosaccharides, PO2- stretching
of external conditions, variations of instruments
symmetric mainly phospholipids, PO2- stretching

112
 Effects of the environmental stress on two fish populations revealed by statistical and spectral analysis

asymmetric mainly phospholipids, C=O stretching mathematical pre treatment. The position of MIR
(symmetric) fatty acids and amino acids, CH2 bending peaks presented in literature, the calculated position
mainly lipids, C=O stretching esther phospholipids, of main absorption bands, and their assignments are
CH2 symmetric stretching in saturated fatty acid presented in Table 1.
chain, CH3 symmetric stretching mainly in proteins, 3.2. FTIR spectra samples after mathematical
C-H stretching of C-H in methine groups, CH2 processing
asymmetric stretching in saturated fatty acid chain,
CH3 asymmetric stretching in saturated fatty acid Lipids play a major role as energy provider for
chain regions infrared bands and the independent organism exposed to stress conditions (Jagadeesan et
variable is etheric extract percent. Regression al., 2005). The infrared spectrum of proteins is
statistics and related plots for different curves of characterized by a set of absorption region named as
estimation regression models were made. A separate amide I (1618-1390 cm-1), amide II (1540-1550 cm-1)
model for each dependent variable was realized. and amide III (1240-1310 cm-1) (Naumann et al.,
2009; Sun, 2010). For muscle and brain before and
3. Results and discussions after extraction, and for extract samples, the protein
bands were founded at 1633-1687 cm-1 (Amide I),
3.1. Spectral band assignments of brain and muscle 1541-1545 (Amide II) and 1278-1312 cm-1 (Amide
sample III). Generally, the muscle tissue (Fig. 1) stores
energy rich molecules like glycogen which is a
FTIR spectroscopy was applied to monitor the glucose polymer. In our study, the glycogen
structural changes in lipids and in amino acids of absorption due to C-O, C-C, C-O-H, C-O-C
proteins (Iwaki et al., 2003), to study the damage and deformation of carbohydrates was assigned to 1028-
coordination effects of heavy metals on photo system 1045 cm-1 region.
II proteins (Nahar and Tajmir-Riahi, 1996), to analyze
the role of the metallothioneins (MT) (Coyle et al.,
2002; Formigari et al., 2007; Vergani et al., 2005), to
study the enzymatic activities in the detoxification
mechanism in response to the presence of different
pollutants in the aquatic environment (Henczová et
al., 2008). Other uses of FTIR spectroscopy were to
detect abnormal changes in cells and tissues with
minimal preparation under in vivo conditions during
minimal invasive investigation (Krafft et al., 2004).
Traditional lipid and fatty acid analyses
require relatively large amounts of biomass (>1 g of
dry biomass), it takes a large amount of time and it is
not recommended for the analysis of a large number
of samples. A high number of variables can have a Fig. 1. FTIR spectra of Oituz chub muscle before extraction
potential impact on the lipid content and composition (solid line) and Mures chub muscle
of the fish, influencing the shape of IR absorption before extraction (dot line)
spectra. For fish IR spectra Laurence and Wolfrum
(2011) describes three distinct absorption bands The zones with an important lipid contribution
which are mainly characteristic for lipids. Two of for our study were assigned at 1052-1240 cm-1, 1718-
them: the CH3/CH2 group is identified from 3025 to 1740 cm-1 and 2853-2963 cm-1. The cellular
2954 cm−1 and the C=O esters group is assigned from components are usually not independent, but interact
1746 to 1654 cm−1 in middle infrared region (MIR). with each other. Such examples are the lipoproteins
The hydroxyl and phosphate groups from the and the biological membranes. Therefore,
phospholipids, the last band, can be found in 1200– mathematics treatment of spectral data is required to
500 cm−1 regions. Arrondo and Goni, 1998 shows that separate the contributions of each cellular component.
phospholipids’ infrared band in the spectrum splits In order to highlight the possible differences at
into three different absorption bands corresponding to molecular level between the two fish populations
the polar (1000-1300 cm-1), interfacial (1700-1750 provided by variations of ecosystems, the infrared
cm-1) and hydrophobic (2800-3100 cm-1) moieties of spectra of these samples were analyzed in detail,
the lipid. using five regions of interest. The regions were
In our previous studies (Lazar et al., 2002; deconvoluted in individual peaks, using information
Lazar et al., 2008) a detailed deconvolution of C-H from Table 1. Figs. 2 and 3 show obtained results
simple and double bonds in hydrophobic region is from multi-peak analysis for four of these regions.
presented. In the present study, the complex spectrum Comparing the two samples (Mures and Oituz
which contains several bands arising from chub muscle before extraction), significant
contributions of various functional groups belonging differences in regions 1000-1150 cm-1 (Figs. 2a and
to carbohydrates, proteins and lipids was analyzed. 3a), 1500-1800 cm-1 (Figs. 2c and 3c) and 2800-3000
For accurate data, all spectra were undergoing to cm-1 (Figs. 2d and 3d) can be observed.

113
 Lazar et al./Environmental Engineering and Management Journal 11 (2012), 1, 109-124

Table 1. FTIR spectral band assignments of brain and muscle samples before, after extraction and for extract samples

Wave number (cm-1) Wave number (cm-1)

Definition of the spectral assignments* References
in literature in our study
2955 Arrondo and Goni, 1998
2957 CH3 str., asym. Venkataramana, 2010
2950-2963
2958 Saturated fatty acids Sun, 2010
2961 Naumann et al., 2009
2924 Arrondo et al., 1998
2927 CH2 str., asym. Venkataramana, 2010
2918-2927
2920 Saturated fatty acids Naumann et al., 2009
2918 Palaniappan et al., 2010
~2898 2899-2908 C-H str. of C-H in methine groups Naumann et al., 2009
Arrondo and Goni, 1998
2871 Naumann et al., 2009
CH3 str., sym.
2870 2865-2882 Movasaghi et al., 2008
mainly proteins
2872 Palaniappan and
Vijayasundaram, 2008
Naumann et al., 2009
2854 -C-H str., sym. of >CH2 in fatty acid chains
Arrondo et al., 1998
2850 2853 - CH2 str. ,sym. methylene chains in membrane lipids
Venkataramana, 2010
2853 Saturated fatty acids
Movasaghi et al., 2008
Arrondo and Goni, 1998
Sun, 2010
1732 C=O str.,
1718-1740 Klaypradit et al., 2010
1753 Esters, phospholipids
Naumann et al., 2009
Dreissig et al., 2009
Naumann et al., 2009
1618-1690 1633-1687 Amide I of proteins
Movasaghi et al., 2008
Movasaghi et al., 2008
1540-1550 1541-1545 Amide II of proteins Palaniappan et al., 2010
Naumann et al., 2009
1450 (control) and 1458
Palaniappan et al., 2010
Zn exposed
1453-1456 CH2 δ, mainly lipids with little contributions from protein Arrondo and Goni, 1998
1456
Venkataramana, 2010
1451
Naumann et al., 2009
1388-1400 1395 C=O str., sym. of COO- : fatty acids and amino acids Venkataramana, 2010
Movasaghi et al., 2008
1240-1310 1278-1312 C-N amide III band components of proteins Naumann et al., 2009
Arrondo and Goni, 1998
PO-2 str., asym.
1220-1250 1225-1240 Naumann et al., 2009
phosphodiesters groups from phospholipids
Dreissig et al., 2009
1170 -CO-O-C asym., str.: mainly nucleic acids Wolkers, 2009
1173 -ester C–O–C str., asym. Arrondo and Goni, 1998
1166-1169
1171 -C–O ester str., single bond Sun, 2010
1176 mainly phospholipids Dreissig et al., 2009
1088 1088 PO-2 str., sym. mainly phospholipids Wolkers, 2009
Arrondo and Goni, 1998
PO-2 str., sym. mainly collagen & phosphodiester groups Movasaghi et al., 2008
1080-1084 1079-1084
from phospholipids Venkataramana, 2010
Dreissig et al., 2009
- Phosphate ester PO2 str. and
1055 1052-1061 Yoshida et al., 1997
C-OH str. of oligosaccharides
-Glycogen band (due to OH str. coupled with δ)
Movasaghi et al., 2008
1044-1045 1040-1045 -C-O str.
Venkataramana, 2010
-C-O δ of C-OH groups of carbohydrates
1035 1035 C-O, C-C str., C-O-H, C-O-C def. (of carbohydrates) Naumann et al., 2009
-Glycogen and CH2OH vib. - C-O str.
1030 1030 Movasaghi et al., 2008
-Collagen & phosphodiester groups of nucleic acids
Glycogen absorption due to C-O and C-C str. and C-O-H
1028 1027 Movasaghi et al., 2008
def.
1020-1022 1020-1022 Glycogen Movasaghi et al., 2008
Str. sym. of dianionic phosphate monoesters of
Movasaghi et al., 2008
970 970 phosphorylated proteins or cellular nucleic acids
Dreissig et al., 2009
-N+–(CH3)3
* δ—bending; str: stretching; vib: vibration; sym: symmetric; asym: asymmetric; def: deformation.

114
 Effects of the environmental stress on two fish populations revealed by statistical and spectral analysis

a) Multi-peak analysis for 950-1150 cm-1 region b) Multi-peak analysis for 1150-1350 cm-1 region

c) Multi-peak analysis for 1450-1800 cm-1 region d) Multi-peak analysis for 2800-3000 cm-1 region

Fig. 2. Multi-peak analysis for four zones (a- 950-1150 cm-1 ; b-1150-1350 cm-1 ; c- 1450-1800 cm-1; d- 2800-3000 cm-1)
of FTIR spectra of Mures chub muscle before extraction

a) Multi-peak analysis for 950-1150 cm-1 region b) Multi-peak analysis for 1150-1350 cm-1 region

c) Multi-peak analysis for 1450-1800 cm-1 region d) Multi-peak analysis for 2800-3000 cm-1 region

Fig. 3. Multi-peak analysis for four zones (a- 950-1150 cm-1 ; b-1150-1350 cm-1 ; c- 1450-1800 cm-1; d- 2800-3000 cm-1)
of FTIR spectra of Oituz chub muscle before extraction

115
 Lazar et al./Environmental Engineering and Management Journal 11 (2012), 1, 109-124

So, these differences between two fish A comparison of the absorption spectrum of
populations are mainly attributed to lipid bands, in Oituz and Mures chub brain etheric extract revealed
concordance with studies of Mecozzi et al. (2009). the bathochromic shift for Mures samples.

3.3. Ultraviolet analysis results

The UV absorption spectra were recorded for

wavelengths from 200 to 400 nm for four samples
from the two studied fish populations: muscle etheric
extract from Oituz River fishes, muscle etheric extract
from Mures River fishes, brain etheric extract from
Mures River fishes, brain etheric extract from Oituz
River fishes. For highlighting the differences, two
samples UV spectra are presented.
In Fig. 4, the spectra for Mures and Oituz chub
brain extract are presented on the same graph, to
highlight the differences, and in Fig. 5 the same
spectra are presented with the appropriate
decompositions.
The appearance of several absorption peaks for
a given chromophore is common for highly
conjugated systems, and is often solvent dependent.
In our case the solvent was diethyl ether. The
information from Fig. 5 are summarized in Table 2.

Fig 5. UV spectra decomposition for Mures (a) and Oituz

(b) chub brain etheric extract

3.4. Comparison between the relative intensity of

major characteristics IR bands

According to spectral band assignments from

Fig 4. UV spectra of Mures and Oituz chub brain etheric Table 1, the functional groups allocated to wave
extract number domains are presented in Table 3.
The mean fish measurements and relative
Table 2. Relative pick intensity of 200-400 nm region for
Oituz and Mures chub brain etheric extract samples
intensity of major characteristic bands are indicated in
Table 4. Because the spectra decomposition was
Mures chub brain made on different regions and not all peaks are
Oituz chub brain etheric extract important for our discussion, the values indicating
etheric extract
Relative relative intensity can be considered expressed in
Wavelength Relative intensity Wavelength
intensity arbitrary units. The fish collected from Mures River is
(nm) (%) (nm)
(%) smaller than the fish collected from Oituz River.
220 6.194 222↑ 14.666 Second, the differences observed between spectra
234 21.279 241↑ 28.0017 presented in Figs. 2 and 3 are confirmed by the results
258 9.1789 presented in Table 4.
263 35.119 273↑ 30.9555 The most important differences between the
289 16.909 301↑ 16.6462 muscle sample’s spectra are in 1052-1061 cm-1
310 1.721 309 0.55012
wavelength domain and, very clear, in 2850-2960 cm-
312 5.1663 1
region. In the case of brain extract, significant
325 12.256
differences appear only for C=O stretching esther
The UV absorption spectra of Mures and Oituz (1700-1750 cm-1), the mainly etheric extract region.
chub brain etheric extract illustrates that conjugation In these conditions we can assume that the differences
of double and triple bonds shifts the absorption between the two samples originating from different
maximum to longer wavelengths. environments can be found mainly in lipid levels
which are majority in etheric extract.

116
 Effects of the environmental stress on two fish populations revealed by statistical and spectral analysis

Table 3. List of FTIR band assignment 3.5. Matrix plot analysis of various infrared spectral
regions
In order to confirm the differences between the
Absorption band (cm-1) Functional group infrared spectra of analyzed samples, Unscrambler X
software was used to obtain the matrix plot for
970 phosphatidylcholine
different samples pairs. The results are presented in
deformation of carbohydrates,
1016-1045 Fig. 6. It can be noticed clear differences between
mainly Glycogen absorption
1052-1061 sphingolipids & oligosaccharides samples before extraction (Figs. 6a, 6b, 6c, 6e) and
PO2- stretching symmetric, mainly between etheric extract samples (Fig. 6f). No
1079-1088 significant differences were observed between Mures
phospholipids
PO2- stretching asymmetric, and Oituz chub brain after extraction (Fig. 6d). This
1225-1240
mainly phospholipids confirms the hypothesis that lipids are the main
C=O stretching symmetric of components of cell membrane correlated with the
1395 COO-, fatty acids and amino environmental characteristics.
acids
CH2 bending, mainly lipids with 3.6. Statistical analysis
1453-1456
little contributions from protein
C=O stretching, ester
1708-1740
phospholipids
3.6.1. Correlation loadings
CH2 symmetric stretching in The regions of interest in our infrared spectral
2853 data were found analyzing the variables using the
saturated fatty acid chains
CH3 symmetric stretching, mainly Principal Component Analysis (PCA) method. The
2865-2871 data matrix of infrared spectra was decomposed into
in proteins
CH2 asymmetric stretching in scores, loadings and residuals. These results indicate
2918-2937
saturated fatty acid chains the degrees of correlation between different FTIR
CH3 asymmetric stretching in spectral regions.
2950-2963
saturated fatty acid chains

Table 4. Measured fish’s mean dimensions and comparison between the relative intensity of major IR characteristic bands

Dimension Oituz samples Mures samples

fish length (cm) 9.1 8.3↑
fish width (cm) 7.4 7↑
fish head length (cm) 2 1.7↑
fish height (cm) 1.7 1.7↔
fish thickness (cm) 0.9 0.8↓
fish weight (g) 8.3 5.3↓
Oituz chub muscle Mures chub muscle Oituz chub brain Mures chub brain
Variable
before extraction before extraction after extraction after extraction
% etheric extracts 6.41 18.98↑ 0 0↔
phosphatidylcholine 0 0 3.6849 2.9142↓
def of carbohydrates 50.5145 68.3398↑ 43.7568 44.7806↑
sphingolipids & oligosaccharides 12.8619 1.274↓↓ 13.1383 14.2865↑
PO2- stretching symmetric mainly 29.2326 26.8084↓ 36.783 33.9356↓
phospholipids
PO2- stretching asymmetric 62.8363 65.334↑ 79.1032 81.1753↑
mainly phospholipids
C=O str (sym) fatty acids and 61.6339 60.1441↓ 54.2814 55.3249↑
amino acids
CH2 bending mainly lipids 38.3661 39.856↑ 45.7186 44.6751↓
C=O stretching esther 3.4023 2.4613↓ 4.3157 1.3134↓↓
phospholipids
CH2 symmetric stretching in 5.6571 9.8285↑↑ 7.1393 6.3138↓
saturated fatty acid chain
CH3 symmetric stretching mainly 13.2695 6.6231↓↓ 1.3267 1.9269↑
in proteins
C-H stretching of C-H in methine 2.5862 21.1388↑↑ 49.583 54.3497↑
groups
CH2 asymmetric stretching in 49.3966 32.3817↓↓ 26.7642 20.6593↓
saturated fatty acid chain
CH3 asymmetric stretching in 29.0907 30.0261↑ 15.1867 16.7503↑
saturated fatty acid chain

117
 Lazar et al./Environmental Engineering and Management Journal 11 (2012), 1, 109-124

a. Oituz Barbus petenyi brain before extraction and Oituz Barbus b. Mures Barbus petenyi muscle before extraction and Oituz
petheny muscle before extraction Barbus petenyi muscle before extraction

c. Mures chub muscle before extraction and Oituz chub before d. Mures chub brain after extraction and Oituz chub brain after
extraction extraction

e. Mures chub muscle before extraction and Oituz chub before f. Mures chub muscle etheric extract and Oituz chub etheric
extraction CH2 CH3 bonds region extract CH2 CH3 bonds region

Fig. 6. Infrared spectra for Mures and Oituz samples

The lowest residual variance is found for 6 the positive correlation between C=O stretching
PCs, but we chose to work with a model consisting of esther phospholipids, CH2 symmetric stretching in
4 PCs. The loading plot, also called map of variables saturated fatty acid chain, CH2 asymmetric stretching
(Fig. 7), displays information about the variables in in saturated fatty acid chain and fingerprint region.
the PCA model. Generally, the variables from the right sided
The plot shows that variables from circle are the variables with the higher degree of
fingerprint region (1000-1300 cm-1) have an extreme correlation. This means that the mentioned variables
position to the right of the plot along PC1, indicating are principal contribution in analyzed infrared
a strong correlation each other. Other observation is spectral data.

118
 Effects of the environmental stress on two fish populations revealed by statistical and spectral analysis

3.6.2. Regression analysis dimensions are closely positive correlated. Also, the
The regression methods establish relations correlations founded from Fig. 7 are confirmed by
between two blocks of variables. First, it was made a this correlated map of variables. The regression
regression model with PLS2 of following groups of coefficients (BW) are used to calculate the response
variables: dimensions of samples and relative value from the infrared spectral relative intensity at
intensity of different infrared peaks (Fig. 8). The different region. The size of the coefficients gives a
predicted versus measured plot displays the sign of which variables have a significant impact
predictions done during calibration. In this case, the assessment on the variables response. The BW
R-square of calibration model was 0.902126. The coefficients are used for interpretation of the
loading weights plot can be studied from the regression model. In this case, the CH2 asymmetric
regression model, to find correlating variables. It is stretching in saturated fatty acid chain and CH3
observed that the dimension of samples (length, asymmetric stretching in saturated fatty acid chain are
width, head length, height, thickness and weight) are relevant for macroscopic properties, like dimensions
strongly negatively correlated with PO2- stretching data (Fig. 10).
symmetric mainly phospholipids, PO2- stretching
asymmetric mainly phospholipids, C=O stretching 3.6.3. Correlation analyzed with SPSS software
(symmetric) fatty acids and amino acids, CH2 bending The Bivariate Correlations from SPSS
mainly lipids corresponds to 1000-1500 cm-1 region software calculates the Pearson's correlation
(Fig. 9). coefficient. The Pearson's correlation coefficient is
Dimensions are correlated with fingerprint significant if it is greater than 0.05.
region (1000-1500 cm-1). As is expected, fish’s

Fig 7. Correlation loadings plot of PCA total samples model

Fig. 8. Predicted versus measured plot for regression model calibration between dimensions of samples and relative intensity of
different infrared peaks

119
 Lazar et al./Environmental Engineering and Management Journal 11 (2012), 1, 109-124

Fig. 9. Correlated map of the variables: dimensions of samples and infrared spectral relative intensity at different region

The variables founded in correlation from

Pearson's analysis will be used in model calibration.
The results of the analysis are presented in the
ANOVA summary table for all correlated variables
(Table 6). The regression plots are represented in
Figs. 11-13.
For each variable, the founded regression
model was linear, with an R square of medium
intensity. The results of linear regression for each
dependent variable are presented in Table 6.

4. Conclusions

The result of chemometric analysis of spectral

Fig. 10. Regression coefficients (BW) from PLS2 methods data is used to take decisions, such as the quality of
in the case of dimensions of samples and infrared spectral environmental ecosystem based on the concentration
relative intensity at different region of each parameter as predicted by computational
analysis of spectra (Brereton, 2007).
Analyzing data from all variables presented in In the scientific literature, the bathochromic
Table 5 can be observed that only C=O stretching shift of a spectral band to lower frequencies observed
esther, CH2 symmetric stretching in saturated fatty in UV/VIS spectrum of etheric extract is associated
acid chain and CH2 asymmetric stretching in saturated with various degrees of lipid oxidation, in close
fatty acid chain are correlated with etheric extract correlation with the environmental factors. It is
percent. These variables are positively correlated known that the oxidative stress should give rise to
which means that when the value of etheric extract modifications of important chromophoric substances
percent increases, the intensity of C=O stretching (Lankmayr et al., 2004). Therefore, the ultraviolet
esther, CH2 symmetric stretching and CH2 spectrum highlights the difference between the fish
asymmetric stretching bands increases. population according to FTIR spectra.
The chemometric analysis shows that the levels of
3.6.4. Regression analyzed with SPSS software lipid components, plays an important role to
The prediction properties were computed from differentiate two fish population from different
the validation variances, residual variances, and the environments confirmed by statistical analyses. So,
leverage of the variable data in the prediction objects. the chemometric analysis result of spectral data is
In addition to the spectral characterization of used to take decisions, such as the quality of
the fish populations, their macroscopic properties environmental ecosystem based on the concentration
were also evaluated, so that these samples could be of each parameter as predicted by computational
used as training sets for a further chemometrical data analysis of spectra (Brereton, 2007).
processing.

120
 Effects of the environmental stress on two fish populations revealed by statistical and spectral analysis

Fig 11. Regression plot between C=O stretching esther and etheric extract percent

Fig 12. Regression plot between CH2 symmetric stretching in saturated fatty acid chain and etheric extract percent

Fig 13. Regression plot between CH2 asymmetric stretching in saturated fatty acid chain and etheric extract percent

121
 Lazar et al./Environmental Engineering and Management Journal 11 (2012), 1, 109-124

Table 5. The Pearson's correlations of etheric extract percent and each functional group from infrared region

Variables Statistical results % etheric extract

Pearson Correlation 1
% etheric extract Sig. (2-tailed)
N 12
Pearson Correlation .083
phosphatidylcholine Sig. (2-tailed) .798
N 12
Pearson Correlation .335
def of carbohydrates Sig. (2-tailed) .287
N 12
Pearson Correlation .195
sphingolipids & oligosaccharides Sig. (2-tailed) .544
N 12
Pearson Correlation .172
PO2- stretching symmetric mainly phospholipids Sig. (2-tailed) .594
N 12
Pearson Correlation .388
PO2- stretching asymmetric mainly
Sig. (2-tailed) .213
phospholipids
N 12
Pearson Correlation .329
C=O str (sym) fatty acids and amino acids Sig. (2-tailed) .296
N 12
Pearson Correlation .457
CH2 bending mainly lipids Sig. (2-tailed) .135
N 12
Pearson Correlation .691*
C=O stretching esther phospholipids Sig. (2-tailed) .013
N 12
Pearson Correlation .681*
CH2 symmetric stretching in saturated fatty
Sig. (2-tailed) .015
acid chain
N 12
Pearson Correlation -.129
CH3 symmetric stretching mainly in proteins Sig. (2-tailed) .689
N 12
Pearson Correlation -.494
C-H stretching of C-H in methine groups Sig. (2-tailed) .102
N 12
Pearson Correlation .628*
CH2 asymmetric stretching in saturated fatty
Sig. (2-tailed) .029
acid chain
N 12
Pearson Correlation -.270
CH3 asymmetric stretching in saturated fatty acid
Sig. (2-tailed) .396
chain
N 12
* Correlation is significant if Pearson's correlation coefficient is greater than 0.05

Table 6. ANOVA summary table for C=O stretching esther, CH2 symmetric stretching in saturated fatty acid chain and CH2
asymmetric stretching in saturated fatty acid chain variables

Model Summary Anova Coefficients

R Adjusted R Std. Error of
Variable R F Sig. t Sig.
Square Square the Estimate
C=O stretching esther .691 .477 .425 1.877 9.135 .013 3.0222 .013
CH2 symmetric
stretching in saturated .681 .464 .411 3.409 8.665 .015 2.344 0.015
fatty acid chain
CH2 asymmetric
stretching in saturated .628 .394 .344 13.717 6.508 .029 2.551 .029
fatty acid chain

In our study, the chemometrical discover the relations between spectral, biological and
characterization of fish population’s sample, the chemical sample properties. Mathematical
Correlation Analysis and the Regression Analysis as pretreatments can contribute to reduce spectral
the main multivariate techniques made it possible to variation (Laurens and Wolfrum, 2011).

122
 Effects of the environmental stress on two fish populations revealed by statistical and spectral analysis

The important spectral regions were assigned Cozzolinoa D., Murraya I., Chreeb A., Scaifec J.R., (2005),
using the regression coefficients, indicating which Multivariate determination of free fatty acids and
infrared regions of the spectra are the main moisture in fish oils by partial least-squares regression
and near-infrared spectroscopy, LWT Food Science and
responsible in calibration model of macroscopic and
Technology, 38, 821-828.
spectral variables. Dreissig I, Machill S., Salzer R., Krafft C., (2009),
The most important differences between the Quantification of brain lipids by FTIR spectroscopy
muscle sample’s spectra from River and Oituz and partial least squares regression, Spectrochimica
populations were attributed to the polar (1000-1300 Acta Part A, 71, 2069-2075.
cm-1) and hydrophobic (2850-2960 cm-1) moieties of Gaetke L.M., Chow C.K., (2003), Copper toxicity,
the lipid components. oxidative stress and antioxidant nutrients, Toxicology,
In the case of brain extract from River and 189, 147-163.
Oituz populations, significant differences appear only Gosav S., Praisler M., Dorohoi D.O., (2007), ANN expert
for interfacial (1700-1750 cm-1) moieties of the lipid system screening for illicit amphetamines using
molecular descriptors, Journal of Molecular Structure,
components. The infrared relative intensity positively 834–836, 188-194.
correlated with etheric extract percent parameter at Formigari A., Irato P., Santon A., (2007), Zinc, antioxidant
1708-1740 cm-1, 2853 cm-1 and 2918-2937 cm-1 was systems and metallothionein in metal mediated-
revealed by statistical analysis of the infrared spectral apoptosis: biochemical and cytochemical aspects,
bands. Note that in the correlation matrix, all infrared Comparative Biochemistry and Physiology Part C:
variables are dependent. Toxicology & Pharmacology, 146, 443-459.
The proposed goals of this study, to use Harja M., Ciocîntă R.C.1, Barbuta M., Lăcrămioara Rusu
attenuated total reflectance (ATR)-FTIR and UV-Vis L., Simion A.I., Bistricianu I.L., (2010), CaCO3
spectroscopy as an alternative method compared with Controllable synthesis by double exchange method
using CaCl2 residual solutions, Environmental
chemical method for evaluating the difference Engineering and Management Journal, 9, 1571-1577.
between two fish population and to apply Henczová M., Deér A.K., Filla A., Komlósi V., Mink J.,
chemometric analysis to predict the etheric extract (2008), Effects of Cu2+ and Pb2+ on different fish
values were achieved. species: Liver cytochrome P450-dependent
In conclusion, the use of spectral information, monooxygenase activities and FTIR spectra,
correlated with statistical methods, allows Comparative Biochemistry and Physiology, Part C,
highlighting the environmental influence on lipid 148, 53-60.
metabolism, for the same fish species from different Hlihor R.M., Gavrilescu M., (2009), Removal of some
hydrographic areas. environmentally relevant heavy metals using low-cost
natural sorbents, Environmental Engineering and
Management Journal, 8, 353-372.
References Jagadeesan G., Kavitha A.V., Subashini J., (2005), FT-IR
Study of the influence of Tribulus terrestris on mercury
Aksu Z., (2005), Application of biosorption for the removal intoxicated mice, Mus musculus liver, Tropical
of organic pollutants: A review, Process Biochemistry, Biomedicine, 22, 15-22.
40, 997-1026. Klaypradit W., Kerdpiboon S., Singh R.K., (2010),
Almeida E.A., Miyamoto S., Bainy A.C.D., Medeiros Application of Artificial Neural Networks to Predict
M.H.G., Mascio P., (2004), Protective effect of the Oxidation of Menhaden Fish Oil Obtained from
phospholipid hydroperoxide glutathione peroxidase Fourier Transform Infrared Spectroscopy Method,
(PHGPx) against lipid peroxidation in mussels Perna Food Bioprocess Technology, 4, 475-480.
perna exposed to different metals, Marine Pollution Krafft C., Stephan B. Sobottka S.B., Schackert G., Salzer
Bulletin, 49, 386-392. R., (2004), Analysis of human brain tissue, brain
Arrondo J.L.R., Goni F.M., (1998), Infrared studies of tumors and tumor cells by infrared spectroscopic
protein-induced perturbation of lipids in lipoproteins mapping, Analyst, 129, 921-925.
and membranes, Chemistry and Physics of Lipids, 96, Lankmayr E., Mocak J., Serdt K., Balla B., Wenz T.,
53-68. Bandoniene D., Gfrerer M., Wagner S., (2004),
Arsene D., Catrinescu C., Drăgoi B., Teodosiu C., (2010), Chemometrical classification of pumpkin seed oils
Catalytic wet hydrogen peroxide oxidation of 4- using UV–Vis, NIR and FTIR spectra, Journal of
chlorophenol over iron-exchanged clays, Biochemical and Biophysical Methods, 61, 95-106.
Environmental Engineering and Management Journal, Laurens L.M.L., Wolfrum E.J., (2011), Feasibility of
9, 7-16. Spectroscopic Characterization of Algal Lipids:
Barth A., Haris P.I. (2009), Biological and Biomedical Chemometric Correlation of NIR and FTIR Spectra
Infrared Spectroscopy, In: Advances in Biomedical with Exogenous Lipids in Algal Biomass, Bioenergy
Spectroscopy, Volume 2, IOS Press BV Nieuwe Research, 4, 22-35.
Hemweg 6B 1013 BG Amsterdam Netherlands. Lazar G., Vascan I., Lazar I., Stamate M., (2002), Some
Brereton R.G., (2007), Biological and Medical Applications properties of sputtered a-C:H, Journal of Non-
of Chemometrics, In: Applied Chemometrics for Crystalline Solids, 299–302, 835-839.
Scientists, John Wiley & Sons Ltd., The Atrium, Lazar G., Bouchet-Fabre B., Zellama K., (2008), Structural
Southern Gate, Chichester, West Sussex PO19 8SQ, properties of nitrogenated amorphous carbon films:
England, 287-317. Influence of deposition temperature and
Coyle P., Philcox J.C., Carey L.C., Rofe A.M., (2002), radiofrequency discharge power, Journal of Applied
Metallothionein: The multipurpose protein, Cellular Physics, 104, 073534, doi.org/10.1063/1.2951622.
and Molecular Life Sciences, 59, 627-647.

123
 Lazar et al./Environmental Engineering and Management Journal 11 (2012), 1, 109-124

Mecozzi M., Pietroletti M., Gallo V., Conti M.E., (2009), Palaniappan PL.RM., Nishanth T., Renju V.B., (2010),
Formation of incubated marine mucilages investigated Bioconcentration of zinc and its effect on the
by FTIR and UV–VIS spectroscopy and supported by biochemical constituents of the gill tissues of Labeo
two-dimensional correlation analysis, Marine rohita: An FT-IR study, Infrared Physics and
Chemistry, 116, 18-35. Technology, 53, 103-111.
Moisescu M.G., Leveque P., Bertrand J.R., Kovacs E., Mir Sun D.W., (2009), Infrared Spectroscopy for Food Quality
L.M., (2008), Microscopic observation of living cells Analyses and Control, 1th Edition, Elsevier Inc.,
during their exposure to modulated electromagnetic Oxford, UK.
fields, Bioelectrochemistry, 74, 9-15. Stefanescu I.A., Gavrila L., Mocanu R.D., Olariu R.I.,
Movasaghia Z., Rehmanb S., ur Rehman I., (2008), Fourier Arsene C., (2011), Bioremediation perspective of
Transform Infrared (FTIR) Spectroscopy of Biological Bacillus Megaterium towards heavy metals in
Tissues, Applied Spectroscopy Reviews, 43, 134-179. environments enriched with phosphogypsum,
Nahar S., Tajmir-Riahi H.A., (1996), Complexation of Chemistry Magazine, 62, 245-249.
heavy-metal cations Hg, Cd, and Pb with proteins of Teodosiu C., Cojocariu C., Musteret C.P., Dascalescu I.G.,
Psii—Evidence for metal–sulfur binding and protein Caraene I., (2009), Assessment of human and natural
conformational transition by FTIR spectroscopy, impacts over water quality in the Prut river basin,
Journal of Colloid and Interface Science, 178, 648- Romania, Environmental Engineering and
656. Management Journal, 8, 1439-1450.
Naumann D., Fabian H., Lasch P., (2009), FTIR Venkataramana G.V., Kumar J. K., Prasad A.G.D., Karimi
Spectroscopy of Cells, Tissues and Body Fluids, In: P., (2010), Fourier Transform Infrared spectroscopic
Advances in Biomedical Spectroscopy, vol. 2, IOS study on liver of freshwater fish Oreochromis
Press BV, Amsterdam, Netherlands. Mossambicus, Romanian Journal of Biophysics, 20,
Neamtu J., Anoaica P.G., (2006), Applications of 315-322.
Electromagnetic Radiations in The Medical Field, (in Vergani L., Grattarola M., Borghi C., Dondero F., Viarengo
Romanian), University of Medicine Publishing House, A., (2005), Fish and molluscan metallothioneins, FEBS
Craiova, Romania. Journal, 272, 6014-6023.
Nicoară M., Ureche D., Plăvan G., Erhan M., (2006), Wolkers W.F., (2009), Biomedical FTIR Spectroscopy of
Comparative analyses of the food spectra of the fish Lipids In: Biological and Biomedical Infrared
populations living in the Rivers Oituz and Casin, Spectroscopy, Barth A. and Haris P.I. (Eds.), IOS
Annals Scientific of the University "Al.I. Cuza" Iaşi, Press, 272-287.
Section I: Animal biology, LII, 87-93. Yoshida S., Miyazaki M., Sakai K., Takeshita M., Yuasa S.,
Palaniappan PL.RM., Vijayasundaram V., (2008), FTIR Sato A., Kobayashi T., Watanabe S., Okuyama H.
study of arsenic induced biochemical changes on the (1997), Fourier Transform Infrared spectroscopic
liver tissues of fresh water fingerlings LABEO analysis of rat brain microsomal membranes modified
ROHITA, Romanian Journal of Biophysics, 18, 135- by dietary fatty acids: possible correlation with altered
144. learning behavior, Biospectroscopy, 3, 281–290.

124

View publication stats