4 mS/cm). Besides, IEX shows poor scalability and is the slowest method

investigated (see Section 2). The use of NPC and ATPS in the purification of betanin produced the worst results. The PEG/(NH4)2SO4 system is based on the differential partitioning of betalains and sugars between the aqueous phases (Chethana et al., 2007 and Neelwarne and Thimmaraju, 2009). Consequently, the presence of large amounts of dextrin in sample C compromises the partitioning, making this approach inefficient. Results reported here are yet to be compared to those obtained using countercurrent chromatography, which has been applied effectively for the purification of betanin and its epimer on the preparative scale (Degenhardt and Winterhalter, 2001, Schliemann and Strack, 1998 and Wybraniec et al., 2010). Betalain

optical isomers can be resolved by several chromatographic methods in the absence of Ibrutinib price chiral stationary phases (Schliemann et al., 1999 and Wybraniec et al., 2009). For example, betanin has a shorter retention time than isobetanin (ΔtR = 0.3 min) Olaparib in the RP-HPLC analysis reported in this work. This difference has been rationalised in terms of a stronger interaction of iBn with the non-polar stationary phase, but no quantitative data is given whatsoever ( Schwartz and Von Elbe, 1980 and Wybraniec, 2005). Therefore, we examined this effect quantitatively through theoretical calculations. Due to the size of betanin, geometry ID-8 optimisation and frequency

calculations were conducted using the semi-empirical method PM6 (Stewart, 2007); however, the energies and the physicochemical properties in water were determined at the SMD/M06-2X/6-311++G(d,p) level (Marenich et al., 2009, Zhao and Truhlar, 2008a and Zhao and Truhlar, 2008b). Furthermore, we limited ourselves to two conformational isomers of both Bn and iBn, namely the axial carboxyl and equatorial carboxyl conformers (Scheme S1, coordinates of optimised geometries are given in the Supplementary data). Due to the lack of structural parameters for betanin and isobetanin, the structure optimisation was verified by comparison of the theoretical electronic transition wavelengths (λTh), determined using the SMD/ZIndo/S approach ( Marenich et al., 2009 and Zerner, 1991), with experimental absorption maxima. The equatorial carboxyl conformers of both Bn and iBn are more planar than the corresponding axial isomers. Results obtained for the equatorial carboxyl conformers of both Bn and iBn are in excellent agreement with experimental data (Table 3). This correspondence is obtained exclusively when fully protonated forms of Bn and iBn are used (pK  a of the carboxylic groups of betanin is 3.4) ( Nilsson, 1970). The absorption maximum is not pH-dependent in the 3–7 pH range ( Azeredo, 2009); however, the results can be rationalised based on the occurrence of hydrogen bonds with strong ion-dipole character between the carboxylate groups and water.

10 environment (Math-Works, Natick, USA) with the PLS-toolbox version 4.0 (Eigenvector Research, Inc., Wenatchee, WA). Samples

(n = 139) were divided into calibration (n = 100) and validation sets (n = 39) by applying the classic Kennard–Stone (KS) selection algorithm to the NIR spectra ( Kennard & Stone, 1969). Calibration and validation sample numbers were, respectively selected at 100 (min. 1.50 g kg−1; max. 82.0 g kg−1; mean 28.6 g kg−1; SD 20.8 g kg−1) and 39 (min. 2.20 g kg−1; max. 68.5 g kg−1; mean 15.8 g kg−1; SD 14.8 g kg−1). Outlier detection was executed to improve model accuracy, remove samples with extreme values, which exhibit increased influence on the model, and eliminate unmodelled residues in the X and Y data responses.

An elliptical joint confidence region (EJCR) was GSK1120212 datasheet calculated to evaluate the ERK inhibitor slope and intercept for the reference regression, and predict values at a 95% confidence interval. Mathematics and applications for the EJCR can be addressed in the following references: González, Herrador, and Asuero (1999), and Goicoechea and Olivieri (2001). All calculations were performed using suitable MATLAB 7.10 approaches. Açaí and palmitero-juçara sample spectra (sample #1 and #80), with respective 3.3 and 36.9 g kg−1 total anthocyanin content (TAC) are presented in Fig. 1. Fig. 1 depicts two spectra, (a) and (b), which characterise the following: a second overtone of the O–H stretch at 982 nm, the first overtone of the O–H stretch

at 1456 nm, and combination bands of the asymmetric and scissor stretch O–H vibrations at 1940 nm. Sugar-related absorption bands were also observed at 926, 1208, 1728, 1762, 2308, and 2348 nm. The spectra were quite homogeneous, and no outliers were identified by a priori visual inspection. The PLS method was initially performed on the entire original spectra wavelength range to develop the NIR model. In this way, TAC could be predicted non-destructively. Noise and systematic behaviour were undesirable features in the spectra, and therefore pre-processing Tacrolimus (FK506) was necessary. The original spectra were subjected to smoothing (first order), multiplicative scattering correction (MSC), and first, second order derivatisation (Savitzky–Golay). Calibration model results for the TAC NIR region in açaí and palmitero-juçara fruits are shown in Table 1. In addition to the PLS models, PLS-SPA, PLS-GA, and iPLS model results are shown. The best performing model results from pre-processing tests (see methods) are provided. F-tests were performed for all models using the prediction set. The results detected no significant differences (95% confidence level) between the best PLS model, the PLS (4) and other models, with the exception of PLS (4) smoothing (3 pts.), PLS (4) first derivative (7 pts.), PLS (4) second derivative (3 pts.

γLW increased slightly (significant, p < 0.01, after 1 h and 168 h of exposure)

when immersed in NaCl + BSA, Table 4, compared with freshly polished and aged coupons. There was no significant difference for γ+. Similar trends were observed for stainless steel exposed to citric acid (pH 2.4), Fig. 4, with increased amounts of released iron and calculated γ− values ( Table 4), and reduced water contact angles with time, Figs. 4a and b. There was also a clear correlation between released amounts of iron and both γ− values and water contact angles, Fig. 4c. The difference of the polar component γ− was significant (p < 0.05) after 24 and 168 h of exposure to citric acid, compared to freshly polished and aged coupons. Corresponding differences for the γLW and the γ+ components were significant (p < 0.05) after 1 and 168 h (γLW), and 168 h (γ+) of exposure in citric DAPT order acid, Table 4. The results imply that all surface energy components increase with time, indicating a surface with increasingly see more amphoteric properties. The results indicate a layer of citrate that becomes more compact and ordered after approximately 24 h of exposure, when low contact angles were observed, approaching conditions for a totally wetted carboxylated surface (<10°) [65]. In all, observed findings indicate that the surface energy increases with time and correlates with released amounts of iron. The objective of this study was

to elucidate the importance and connection between surface physicochemical characteristics including surface energy and wettability and surface oxide composition with the release of iron from stainless steel surfaces in complexing biological media. No correlation was observed between the surface

oxide composition of stainless steel (grade AISI 304) and calculated surface energies or the wettability for polished, aged surfaces in non-complexing solutions. Instead, the surface contamination (adventitious atmospheric carbon or from cleaning Teicoplanin solvents) probably strongly influenced the surface energy of stainless steel. The amount of released iron from stainless steel in solutions containing BSA or citrate (10 mM NaCl + 10 g/L BSA, 5 g/L citric acid) strongly correlated with the measured wettability and calculated surface energy. The surface energy components (γLW, γ+, γ−) increased and the static water contact angles decreased with increased amount of released iron. These observations and the delay in released amounts of iron with time strongly suggest an adsorption-controlled ligand-induced metal release process in the presence of BSA and citrate. The Swedish Research Council (VR), grant number 2013-5621, and Göran Gustafsson’s prize for young researchers (J. Hedberg) are gratefully acknowledged for financial support. “
“Les auteurs demandent de remplacer le mot « résection » par le mot « ovariectomie », à trois reprises, dans le texte de leur article : 1.

A large number of articles/reports concerning levels and time trends of POPs in mothers’ milk from this monitoring program are summarized (with some new original data included) in a review article from the year 2000 ( Norén and Meironyte, 2000), which reports decreasing levels of “dioxins” over time. This is supported by Lignell et al. (2009), who report decreasing levels of POPs, including “dioxins”, in mothers’ milk from Sweden

during 1996–2006. Only a few time series with multiple year sampling of mothers’ milk exists, especially with samples from the last decade. In this study we re-analyze a set of samples from the original (composite) time trend study (Norén and Meironyte, 2000), to test comparability, as well as new samples from 1999 to 2011. This will help find more us answer if the decreasing concentrations of “dioxins” are leveling off, i.e. what is the trend for the first decade of the 21st millennia, and if it is possible to compare the established concentrations from previous studies directly. Hence, the aims of the present study were to assess temporal trends of PCDD, PCDF and DL-PCB in mothers’ milk from Stockholm, 1972–2011, and to compare the results with previous analyses of

some of the older samples. Three major concerns were considered when choosing mothers’ milk samples; i) to ensure comparability between the new and the previous analyses of Swedish mothers’ milk; ii) to add samples taken in the past to fill gaps in the previous time trend and iii) to expand Fulvestrant cost the aforementioned time trend study ending in 1997 (Norén and Meironyte, 2000) to obtain data ranging from 1972 to 2011. In total, 30 samples were analyzed and Lck eight of these were non-identical samples from a given year. The samples consisted of pooled mothers’ milk from multiple donors, all healthy native Swedish, but were not exclusively from primiparae. Further information concerning sample composition is presented in Table 1. The samples, 50 g each, were provided by the Swedish Environmental Specimen Bank, Department of Environmental Research and Monitoring, Swedish Museum of Natural

History. The samples were prepared in house before they were shipped, on dry ice, to Eurofins GfA Lab Service GmbH, Hamburg, Germany, for analysis according to the method described by Reis et al. (2007). In brief the method could be described as follows: 13C-labeled surrogate standards were added to the samples followed by liquid–liquid extraction and subsequent gravimetric lipid determination prior to multiple column clean-up, including carbon column purification. The purified extracts were analyzed by GC/HRMS. To test for significant log-linear trends for PCDDs, PCDFs and DL-PCBs, log-linear regression analysis was performed for the entire investigated time period and for the most recent 10 years using the yearly arithmetic mean values.

1B); 3-yr-old plants had three and four leaves, respectively, each with three to five leaflets (Fig. 1C). Quantitative FT-IR spectra from ginseng leaves of different cultivars (Fig. 2A) and cultivation

ages (Fig. 2B) were obtained (Fig. 2). The most significant spectral variation among the four ginseng cultivars was observed in the polysaccharide region (1,050–1,150 cm−1) and amide region (1,550–1,650 cm−1) of the FT-IR spectra (Fig. 2A). The quantitative spectral variation among cultivation ages of ginseng leaves was also observed in GW-572016 order the polysaccharide region (1,050–1,150 cm−1) and in a broad range (1,200–1,500 cm−1) corresponding to phospholipid/DNA/RNA [39] of the FT-IR spectra (Fig. 2B). These FT-IR spectral variations from leaf samples simply indicate that there were qualitative and quantitative metabolic changes between the cultivars and cultivation ages of ginseng. The PCAs of the FT-IR spectral data are displayed in a two-dimensional plot using the first two principal components (Fig. 3A), which together accounted for 37.5% and 15.7% (53.2% total) of the total variation. PCA score plot showed that most leaf samples belonging GDC 0199 to the same cultivation age segregated into broad boundaries indicating that PCA had a relatively high distinguishing capacity between ginseng leaf samples with a cultivation age dependent

manner. Identifying the most significant spectral variables (i.e., those exhibiting the greatest variance on PC 1 and PC 2 scores) for sample separation is possible using PCA loading values. A PC score loading plot based on Branched chain aminotransferase PCA data from ginseng leaves is displayed in Fig. 3B. Significant FT-IR spectral variables for determining PC 1 and PC 2 were mostly distributed in the polysaccharide region (1,050–1,150 cm−1) and amide region (1,550–1,650 cm−1)

of the FT-IR spectra, respectively (Fig. 3B). These results indicate that qualitative and quantitative metabolic changes corresponding to polysaccharides and protein/amide regions I and II were important variations related to cultivation age. PLS-DA also indicated that a more discrete clustering pattern of ginseng leaves was possible (Fig. 4A). Most samples belonging to the same cultivation age, except the 2-yr-old open-pollinated variety, were grouped more closely in discrete clusters than they were in the PCA, indicating that PLS-DA was more clearly able to distinguish between cultivation ages. A dendrogram based on PLS-DA of the FT-IR spectral data (Fig. 4B) showed that the 12 categories of ginseng cultivars were separated into two major groups in a cultivation age-dependent manner without the 2-yr-old open-pollinated variety. The first group consisted of all 1-yr-old ginseng cultivars and the 2-yr-old open-pollinated variety.

However, since VIF neither detects multiple near-singularities nor identifies the source of singularities (Rawlings et al., 2001), condition index (CI) was evaluated for all variables within the models. This index is the square root

of the ratio of the largest eigenvalue to the corresponding eigenvalue from the matrix. Similar to VIF, the CI indicates weak dependencies when 10 > CI > 30 to high dependencies when CI > 30. Additional data to test the models were not available, thus cross-validation analysis was performed using the predicted residual sum of the squares (PRESS) statistics (Allen, 1971), which is the sum of squares of the difference between each observation find more and its prediction when that observation was not used in the prediction equation. The root mean square error from the cross validation analysis (CV-RMSE) was then calculated as the square root of the ratio between the PRESS statistic and the number of observations. The CV-RMSE is an indicator of the predictive power of the model, thus a small CV-RMSE is desirable. The significance level used for all

the statistical tests was α = 0.05 (p-value < 0.05). This p-value was used to evaluate if the variables included in the model were statistically significant as well. The squared semipartial check details correlation coefficients (SSCC) were calculated using partial sum of squares to determine the contribution from each variable to the models, while controlling the effects of other independent variables within the model. These coefficients represent the proportion of the variance from the dependent variable associated uniquely with much the independent variable. Stand age ranged from 11- to 26-year-old. Forest canopy was closed in all plots, except

for the plots in NSD that had the spacing twice as large as that traditionally used in forest operations, and the plots from RW18 that were thinned. Table 2 summarizes the average growth metrics of plots, within the study sites, as treatment and control, and in the case of NSD, these were distinguished by the number of trees per hectare. In RW19 all plots were classified as fertilized, since the stand had been under traditional forest management. Studies in which there were different levels of fertilization were classified together as fertilized, regardless of the rate and frequency of nutrient additions. In RW18, thinning was recently applied to some of the control and fertilized plots, thus the plots at this site were also classified by the number of trees per hectare. Individual tree height ranged from 4.8 to 27.9 m and averaged 15.7 m among all the study areas, the highest standard deviation (>2 m) from the mean of tree height was observed in the SETRES and Henderson studies. Crown length ranged between 0.8 (a damaged tree) and 10.8 m, and averaged 6.9 m.

, 2004). One of the mutations detected by these authors in the G protein after 33 NMSO3

passages (i.e., F168S) was also detected by us in the PG545 resistant RSV. In conclusion, PG545 is a novel inhibitor of RSV that exhibits virucidal selleck kinase inhibitor activity, and the development of RSV resistance to this compound is slow. Given the fact that PG545 and related glycosides also exhibited potent virucidal activities against HSV (Ekblad et al., 2010) and HIV (Said et al., 2010), the cholestanyl-conjugated oligosaccharides are promising lead compounds for treatment and/or prevention of infections caused by viruses that use GAGs as initial receptors. This work was supported by grants from the Torsten and Ragnar Södeberg Foundation, the Mizutani Foundation for Glycoscience, the Swedish

Research Council, and the Sahlgren’s University Hospital Läkarutbildningsavtal. C.R. Andrighetti-Fröhner was supported by the graduate research fellowships from CNPq MCT, Brazil. “
“This article has been retracted: please see Elsevier Policy on Article Withdrawal (http://www.elsevier.com/locate/withdrawalpolicy). Reason: The Editorial Board of the Journal of Endodontics has received an allegation of data fabrication of an article published in the February issue of the Journal (Yang et al., JOE 38:170–76, 2012). After review of available evidence, the Editorial Board believes there is sufficient evidence to doubt the veracity of the publication. Accordingly, we have retracted these publication of this article. “
“In the article “Effect of Dexamethasone on Root Resorption after Delayed Replantation of Rat Tooth” (J Endodod 2003;29[12]:810–3) click here by Kee-Yeon Keum, Oh-Taik Kwon, Larz S. Spångberg, Chong-Kwan Kim, Jin Kim, Moon-Il Cho, and Seung-Jong Lee, the first author’s name is misspelled. It should be Kee-Yeon Kum. “
“The article “Postoperative Pain after Manual and Mechanical Glide Path: A Randomized Clinical Trial” by Damiano Pasqualini, Livio Mollo, Nicola Scotti, Giuseppe Cantatore, Arnaldo Castellucci, Giuseppe Migliaretti, and Elio Berutti (J Endod 2012;38[1]:32–6) should have included this statement in the author information

section: “Giuseppe Cantatore, Arnaldo Castellucci, and Elio Berutti declare that they have financial involvement (patent licensing arrangements) with Dentsply Maillefer with direct financial interest in the materials (PathFile) discussed in this article. “
“Oliver Pontius and Jeffrey W. Hutter, authors of the article “Survival Rate and Fracture Strength of Incisors Restored with Different Post and Core Systems and Endodontically Treated Incisors without Coronoradicular Reinforcement” (J Endod 2002;28[10]:710–5), would like to add 3 coauthors to their article. The author order now reads: Pontius O, Nathanson D, Giordano R, Schilder H, Hutter JW. They would also like to add the following citation to their article: Strub JO, Pontius O, Koutayas S.

coli DH10Bac for the construction of recombinant bacmids. These bacmids, containing the sequence of the protein with antiviral activity and other chosen proteins, were used for the expression of the proteins in a baculovirus/Sf9 cells system. Three passages of the recombinant virus were performed in Sf9 cell cultures so far. At the moment, titers of the baculovirus obtained

in the different passages as well as the antiviral activity of the recombinant protein produced in this system were determined. To eliminate the possibility that the observed effect is due to characteristics of the www.selleckchem.com/products/azd5363.html construct other than the antiviral activity itself, we used the same approach and procedures to construct recombinant bacmids expressing other L. obliqua proteins, namely LOH-19 and 8-LOH ( Veiga et al., 2005). These two recombinant bacmids, as well as an empty bacmid were used Selleck AZD0530 to treat Sf-9 cells infected with a picornavirus. The results showed that the empty bacmid or those expressing the other recombinant proteins were not effective in inhibiting the replication of EMC

virus, presenting results similar to those of the control of infected cells and of the untreated cells. On the other hand, when infected cultures were treated with the recombinant antiviral, there was a reduction of about 3 logs in the viral titers in comparison to that of controls. Therefore, when the purified antiviral protein was used, the reduction in virus produced was around 4 logs, showing that the recombinant antiviral protein remained fully

active ( Table 1). We are currently testing the effect of the antiviral purified recombinant protein on enveloped viruses (measles, rubella and herpes simplex). Preliminary data have shown that the purified recombinant protein is able to reduce by at least 4 logs the replication of the rubella virus and by about 6 logs the replication of the herpes simplex virus (data not shown). To facilitate purification, a His-tag sequence was included in the C-terminal region of the proteins rAVLO, LOH-19-AY829833 and 8-LOH. The protein was separated by SDS–PAGE and transferred to nitrocellulose membranes (Sambrook and Russell, 2001). After transfer, the membrane was marked with the anti-histidine antibody to confirm the presence of the protein. The result is shown in Fig. 3. As can be seen, there was the presence of a band with Idoxuridine strong labeling with the antibody, demonstrating the expression of the antiviral protein. Viral diseases affect hundreds of millions of people worldwide every year. Even though some antiviral drugs are under clinical trials, 50% of them are directed toward the treatment of HIV. Therefore, there is a need for the development of antiviral agents specific for emerging newly-recognized human pathogens (such as SARS coronavirus and influenza viruses H5N1 and H1N1) (Delcroix and Riley, 2011). Recently, various studies have reported the antiviral properties in products obtained from arthropods.

, 2008, Guyenet et al., 2010 and Nattie and Li, 2009). The retrotrapezoid nucleus (RTN), locus coeruleus, medullary raphe, hypothalamic orexinergic neurons and the NTS neurons are the main sites suggested to be involved with the central chemoreception (Abbott et al., 2009, Biancardi et al., 2008, Dean et al., 1989, Deng et al., 2007, Johnson et al., 2008, Moreira et al., 2007, Mulkey et al., 2004, Nattie

and Li, 2008, Richerson, 2004, Takakura et al., 2006 and Williams et al., 2007). The main focus of the present study is to reexamine the question whether the NTS, particularly the commissural buy Pictilisib division of the NTS caudal to the area postrema, is involved in chemoreception. Studies from the literature have suggested that acid-responsive neurons are located in the NTS and the acidification of the NTS region alters breathing (Dean et al., 1989 and Nattie and Li, 2008). Additionally, previous studies have tested the effects of the lesions or the glutamatergic blockade in the commNTS, suggesting Selleck Bcl-2 inhibitor that this region is essential for the cardiorespiratory responses to the peripheral chemoreceptor activation (Colombari et al., 1996, Sapru, 1996 and Braga

et al., 2007). On the other hand, a more recent study evaluated the effects of muscimol microdialysis in a more rostral portion of the commNTS, suggesting that the rostral portion of the commNTS is involved mainly with respiratory responses to hypercapnia (Nattie and Li, 2008). Based on the assumptions described above, in the present

study, also using muscimol injection to block the neuronal activity, we investigated the importance of the neurons located in a more caudal portion of the commNTS for the cardiorespiratory responses elicited by chemoreflex activation with hypoxia (8–10% O2 in the breathing air) or hypercapnia (8–10% Hydroxychloroquine purchase CO2 in the breathing air) in conscious or anesthetized rats. The experiments were performed on 36 male Holtzman rats weighing 300–330 g. The animals were housed individually in stainless steel cages in a room with controlled temperature (24 ± 2° C) and humidity (55 ± 10%). Lights were on from 7:00 am to 7:00 pm. Standard Guabi rat chow (Paulinia, SP, Brazil) and tap water were available ad libitum. The experimental protocols were approved by the Animal Experimentation Ethics Committee of the Institute of Biomedical Science of University of São Paulo. All efforts were made to minimize animal discomfort and the number of animals used. Rats were anesthetized with intraperitoneal (i.p.) injection of ketamine (80 mg/kg of body wt) combined with xylazine (7 mg/kg of body wt) and placed in a stereotaxic frame (model 1760; David Kopf Instruments). A stainless steel cannula was implanted into the commNTS using the coordinates: 15.0 mm caudal to bregma, in the midline and 7.5 mm below dura mater.

g., Maya, Turner and Sabloff, 2012; Chaco Canyon, English MEK inhibitor et al., 2001; Near East; Artzy and Hillel, 1988 and Jacobsen and Adams, 1958). There are also success stories indicating both environmental and sociopolitical resilience and adaptation in the face of environmental change (McAnany and Yoffee, 2010; Luzzadder-Beach et al., 2012 and Butzer, 2012). The collapse or persistence of ancient states in the context of unintended anthropogenic environmental change therefore provides a starting point for studying the complex socio-ecological

dynamics promoting societal sustainability or collapse under changing conditions (Butzer, 2012). The complexity

of these interactions provides lessons for policy makers considering anthropogenic global climate change today. The staggered and widespread collapse of Classic Maya political centers between AD 750 and 1000 provides a case in point. More than 113 monument-bearing low density urban centers emerged in the tropical lowlands at different times during the Classic Period; each with populations ranging from ∼10,000 (e.g., Uxbenka; Prufer et al., 2011 and Culleton, 2012) to 60,000 plus (e.g., Tikal, Culbert and Rice, MK-2206 1990) people. In addition, thousands of smaller sites, many dating to this interval, dotted the landscape between these larger enough population centers (Witschey and Brown, 2013). It is difficult, if not impossible, to estimate how many people were living in the tropical Maya lowlands, but estimates range between three (Culbert and Rice, 1990) and 10 million at AD 700 (Scarborough and Burnside, 2010). Stone monuments at ∼35 primary political centers during the Late Classic

Period (AD 600–900) show a complex network of antagonistic, diplomatic, subordinate and kinship relationships (Munson and Macri, 2009). The collapse of Classic Maya political systems played out over centuries starting with the first evidence for political fragmentation in the Petexbatun region between AD 760 and 800 (Demarest, 2004a, O’Mansky and Dunning, 2004 and Tourtellot and González, 2004). A 50% reduction in the number of centers with dated-stone monuments between AD 800 and 825 signaled the widespread collapse of kingship and this important political institution had largely disappeared in the central and southern lowlands by AD 900. Politically important centers shifted north to the Yucatan as centers failed in the southern and central Maya lowlands (Sabloff, 2007), and depopulation took centuries and involved migration, reorganization, and persistence in some regions (Laporte, 2004 and Webster et al., 2004).