Magnetic Fields & Gene Expression in Yeast: Study on Protein Synthesis & Cell Processes, Notas de estudo de Engenharia de Produção

Baixe Magnetic Fields & Gene Expression in Yeast: Study on Protein Synthesis & Cell Processes e outras Notas de estudo em PDF para Engenharia de Produção, somente na Docsity! 25 RADIATION RESEARCH 160, 25–37 (2003) 0033-7587/03 $5.00 q 2003 by Radiation Research Society. All rights of reproduction in any form reserved. Effect of Power-Frequency Magnetic Fields on Genome-Scale Gene Expression in Saccharomyces cerevisiae Satoshi Nakasono,a,b,1 Craig Laramee,b Hiroshi Saikia and Kenneth J. McLeodb a Bio-Science Department, Abiko Research Laboratory, Central Research Institute of Electric Power Industry, 1646 Abiko, Abiko-City, Chiba 270- 1194, Japan; and b Program in Biomedical Engineering, State University of New York at Stony Brook, Stony Brook, New York 11794 Nakasono, S., Laramee, C., Saiki, H. and McLeod, K. J. Effect of Power-Frequency Magnetic Fields on Genome-Scale Gene Expression in Saccharomyces cerevisiae. Radiat. Res. 160, 25–37 (2003). To estimate the effect of 50 Hz magnetic-field exposure on genome-wide gene expression, the yeast Saccharomyces cer- evisiae was used as a model for eukaryotes. 2D PAGE (about 1,000 spots) for protein and cDNA microarray (about 5,900 genes) analysis for mRNA were performed. The cells were exposed to 50 Hz vertical magnetic fields at 10, 150 or 300 mT r.m.s. for 24 h. As positive controls, the cells were exposed to aerobic conditions, heat (408C) or minimal medium. The 2D PAGE and microarray analyses for the positive controls showed high-confidence differential expression of many genes including those for known or unknown proteins and mRNAs. For magnetic-field exposure, no high-confidence changes in expression were observed for proteins or genes that were re- lated to heat-shock response, DNA repair, respiration, protein synthesis and the cell cycle. Principal component analysis showed no statistically significant difference in principal com- ponents, with only insignificant differences between the mag- netic-field intensities studied. In contrast, the principal com- ponents for the positive controls were significantly different. The results indicate that a 50 Hz magnetic field below 300 mT did not act as a general stress factor like heat shock or DNA damage, as had been reported previously by others. This study failed to find a plausible differential gene expression that would point to a possible mechanism of an effect of mag- netic fields. The findings provide no evidence that the mag- netic-field exposure alters the fundamental mechanism of translation and transcription in eukaryotic cells. q 2003 by Ra- diation Research Society INTRODUCTION Public concern regarding the health effects of power-fre- quency electric and magnetic fields (EMFs) has been in- creasing since an epidemiological study in 1979 reported 1 Address for correspondence: Bio-Science Department, Abiko Re- search Laboratory, Central Research Institute of Electric Power Industry, 1646 Abiko, Abiko-City, Chiba 270-1194, Japan; e-mail: nakasono@ criepi.denken.or.jp. an increased cancer risk in children living near power lines (1). Since then, many studies have been conducted in an effort to explain the findings, and some recent research sug- gests that weak power-frequency fields (50/60 Hz) may ex- ert biological effects. The NIEHS working group report (2) of the EMF Research and Public Information Dissemination Program stated that extremely low-frequency (ELF) EMFs may be carcinogenic to humans. However, a mechanism for the carcinogenic or biological effects of EMFs has not been forthcoming, and most laboratory results are in disagree- ment with this conclusion (2–6). Several investigators have reported observations of an effect of weak magnetic fields (below 1 mT) on gene ex- pression in bacteria (7) and on heat-shock proteins (HSPs) in yeast (8) and mammalian cells (9, 10). Additional reports suggest that the pattern of synthesis is similar to the re- sponse to general environmental stress (11). This type of response with induction of HSPs is widely reported in bac- teria, yeast and human cells. In contrast, other reports in- dicate no such response in bacteria (12) and human cell lines (13). Some reports suggest that these discrepancies could arise from differences in exposure systems, differ- ences in cell karyotypes between the research groups (2, 5), or a misunderstanding of experimental noise (12). McCann and her colleagues reviewed the literature on the genotoxic potential of EMFs in 1993 and again in 1998 (3, 4). Most of the reports reviewed showed that ELF EMFs ranging from 0.15–5 mT were not genotoxic. Nevertheless, some reports were positive. An exposure to a 50 Hz, 400 mT magnetic field increased the mutation frequency in MeWo cells (14), and a 60 Hz, 0.1–0.5 mT magnetic field inhibited DNA repair in rat brain cells (15, 16) in a reaction that could be blocked by free radical scavengers, suggesting involvement of oxygen radicals. The general stress response (elevated temperature, tox- ins, heavy metals, free radicals, etc.) is found in all bacteria, plant and animal cells, and it is remarkably conserved throughout evolution. The mechanism of the stress response has been widely investigated in the yeast Saccharomyces cerevisiae, whose genome has recently been fully se- quenced and which provides a useful model system for studies of eukaryotes. Recently databases for yeast have 26 NAKASONO ET AL. FIG. 1. Comparison of 2D image of yeast total protein between control and positive controls. 7: increased level of protein synthesis compared to control gel. N: decreased level of protein synthesis. The change in synthesis levels was estimated by visual observation. Protein; 1002125 mg. Control conditions: for 24 h at 258C, YPD medium. Aerobic conditions: for 12 h at 258C with 1 liter air/min, YPD medium. Heat shock: at 508C for 1 h, YPD medium. Minimal medium: for 24 h at 258C, minimal medium. Three independent experiments (control and exposure) were done for each condition. The difference in the protein profiles was estimated visually in the same experimental set. The gel images shown are typical gels for each condition. 29POWER-FREQUENCY MAGNETIC FIELDS AND GENE EXPRESSION IN YEAST TABLE 1 Extended Heat shock Protein mRNA ratio CL Minimal Protein mRNA ratio CL MF Protein 10 mT mRNA ratio CL 150 mT mRNA ratio CL 300 mT mRNA ratio CL — NI — UP UP 0.4 2.8 0.6 3.4 3.4 C C (UP) — A (UP) A (UP) — — — — — 0.8 ND 0.6 1.2 1.5 — — — — — — — — — — 1.2 0.9 ND 1.2 1.2 — — — — — 1.0 0.7 1.1 1.1 0.9 — — — — — 1.1 0.6 0.8 1.1 1.0 — — — — — ND — — NI NI 1.8 0.3 0.4 0.2 0.3 — D (DOWN) C (DOWN) A (DOWN) C (DOWN) ND — — NI NI 1.2 0.9 1.3 0.6 0.8 — — — — — ND — — NI NI 1.0 1.1 0.8 0.6 1.0 — — — — — 0.8 1.0 1.4 1.2 0.9 — — — — — 1.0 1.3 0.9 1.2 1.1 — — — — — — — NI NI — 0.7 0.3 0.6 0.2 0.4 — C (DOWN) — A (DOWN) C (DOWN) — — NI NI — 1.1 1.2 2.0 1.6 2.3 — — C (UP) — C (UP) — — NI NI — 0.8 0.9 0.7 1.2 1.1 — — — — — 1.2 0.9 1.2 1.1 0.8 — — — — — 1.0 0.8 1.1 1.2 1.0 — — — — — NI — NI — NI 2.4 1.3 1.3 0.8 0.3 C (UP) — NI — A (DOWN) NI — UP — NI 0.9 1.5 2.3 0.5 0.9 — — E (UP) — — — — NI — NI 0.9 0.7 0.8 0.9 0.9 — — — — — 0.8 0.9 1.0 1.4 1.2 — — — — — 0.9 1.1 1.0 1.1 1.2 — — — — — UP — — NI — 4.0 1.0 2.1 0.4 1.3 A (UP) — C (UP) C (DOWN) — UP — — NI — 4.8 1.6 0.6 1.5 1.7 A (UP) — — — — — — — ND — 1.4 1.5 0.7 1.0 1.2 — — — — — 0.5 ND 1.1 1.4 0.7 — — — — — 0.9 2.1 1.2 1.1 1.1 — C (UP) — — — — NI NI NI NI 0.3 0.7 — 1.0 1.5 A (DOWN) — — — — — NI NI NI — 1.0 1.1 1.1 0.8 2.3 — — — — C (UP) — NI NI NI — 0.6 0.7 1.0 0.6 0.9 — — — — — 1.0 0.8 0.9 1.0 0.6 — — — — — 1.0 1.0 1.1 0.9 1.6 — — — — — UP NI — NI — 3.2 4.1 ND 7.2 1.5 A (UP) A (UP) — A (UP) — — NI — NI — 1.5 0.6 1.0 1.0 1.0 — — — — — — NI — NI — 1.4 0.9 0.9 1.0 1.2 — — — — — 1.3 1.0 ND 0.6 1.0 — — — — — 1.5 0.8 1.1 0.9 1.2 — — — — — NI NI — NI NI 0.4 4.6 0.4 0.5 0.7 D (DOWN) A (UP) C — — NI NI — NI NI 1.3 1.2 0.9 1.2 0.7 — — — — — NI NI — NI NI 1.1 1.5 1.3 0.9 0.8 — — — — — 0.7 1.0 1.0 1.0 1.0 — — — — — 0.9 1.0 0.8 0.9 0.9 — — — — — NI NI — NI NI 0.6 ND 0.9 1.5 4.5 — — — — A (UP) NI NI — NI — 0.9 2.0 0.7 ND ND — E (UP) — — — NI NI — NI — 1.3 0.8 0.7 0.4 0.8 — — — D (DOWN) — 1.2 0.6 1.0 ND 0.7 — — — — — 1.2 1.1 1.2 0.5 0.8 — — — — — — NI — NI NI 0.7 0.4 1.3 0.3 ND — D (DOWN) — A (DOWN) — — NI — NI NI 0.8 1.5 1.3 1.0 ND — — — — — — ND — NI NI 0.9 1.2 0.7 1.0 0.5 — — — — — 1.0 ND 1.1 0.9 1.5 — — — — — 0.9 1.1 0.9 1.1 1.2 — — — — — 30 NAKASONO ET AL. TABLE 1 Continued ORF Gene name Protein name on 2D gel Process Function Aerobic Protein mRNA ratio CL YBR267W YDR099W YER177W YKL056C — BMH2 BMH1 YB9M BMH2 BMH1 TCTP unknown unknown unknown unknown unknown suppresses clathrin deficiency unknown; similar to mammalian 14-3-3 proteins unknown NI — — NI 1.1 1.0 1.3 1.6 — — — — Notes. UP; up-regulation in exposure condition, DOWN; down-regulation in exposure condition. —; no change, ND, not detectable, NI; not identified on 2D. CL; confidence level of mRNA ratio, A, B, C . . . ; confidence level from microarray result was estimated as shown in experimental section. A; highest, B; high, C; medium, D; low, E; lowest. were 258C for 24 h without air bubbling. For magnetic-field exposure, the cells were incubated with the sinusoidal 50 Hz, vertical magnetic field at 10, 150 or 300 mT for 24 h at 258C without air bubbling. For the positive control, the cells were incubated as follows: For aerobic condi- tions, the cells were incubated in YPD medium with air bubbling at 1 liter/min for 12 h. For heat shock, the cells were incubated at 408C for 1 h after incubation in YPD medium at 258C for 23 h. For minimal medium, the cells were incubated in minimal medium [1.7 g of YNB- AA/AS, 5 g of (NH4)2SO4, 20 g of dextrose in 1 liter of deionized water] at 258C for 24 h. Magnetic-Field Generator An exposure system (12, 26) capable of producing high-intensity mag- netic fields up to 10 mT r.m.s. on both the vertical and the horizontal axis was used to provide a large uniform magnetic-field environment (40 cm3, field variation below 5%). The stray field from the exposure coils was less than 0.01 mT at the sham-exposure system, which was located 25 m from the center of the exposure system. For the higher-intensity magnetic fields, a second magnetic-field exposure system was used which generated uniform flux densities up to 300 mT. The field variation was below 5% in the 10-cm3 exposure space, and the stray field was less than 0.01 mT at the sham-exposure system. Vibration and heat from the coil were carefully removed by a separated incubator stand and using a jack- eted plastic incubator. Sham-exposure experiments, which used a non-energized coil, were carried out to determine the differences in the culture conditions in the control system and the non-energized exposure system. No differences between the 2D gels were found in the visual comparison. Thus differ- ences in the culture systems can be neglected in these experiments. 2D PAGE Analysis of Protein After the exposure, approximately 40 ml of the cell suspension was kept in ice water for 10 min. The cells were collected by centrifugation at 3,000g for 15 min at 48C. Protein samples were prepared by sonication as described previously (12). The protein concentration was measured by the Bradford method using a protein assay kit (No. 500-0006, Bio-Rad Laboratories). Two-dimensional PAGE was done as described previously (12) using the SWISS-2DPAGE online database to identify proteins.2 A nonlinear immobilized pH gradient gel (3.5–10.0 NL IPG 18 cm, Pharmacia Bio- tech Co.) was used as the first dimension. In second-dimension electro- phoresis, a vertical gradient slab gel was run with the modified Laemmli- SDS discontinuous system (24). Gradient gel was used at acrylamide/ piperazine diacrylyl of 9–16%. Proteins in the 2D gels were stained with an ammoniacal silver staining method2 (12). The gel image was imaged by using PDI Scanning Facility from Pharmacia Biotech. The 2D gels were analyzed by comparing them to the SWISS-2DPAGE database2 by visual observation. Approximately 1,000 proteins were sep- arated by 2D PAGE. Approximately 60 proteins, including some stress- response proteins, were closer for analysis on the 2D gels (data not shown). Protein levels were compared with those on control gels from the same batch (same time, same preculture) by visual observation to avoid false positives. At least three sets of gels from three independent experiments were used to estimate the changes in protein levels. When some differences in the levels of were found on the gels, the reproduc- ibility was checked by using 2D gels from different batches. Typical 2D gels are shown in Figs. 1 and 2. cDNA microarray Analysis of RNA Extraction of total RNA from yeast cells was carried out by the hot phenol method (28). The quality of the RNA sample was checked using gel electrophoresis and absorption measured at 260 and 280 nm. Poly (A)1 RNA was isolated using Oligotex-dT30 (TaKaRa Co.). Only mRNA samples with an A260/A280 ratio of above 1.9 were used for the follow- ing steps. The mRNA samples were labeled with the fluorescent dyes Cy3 and Cy5 for competitive hybridization to the yeast cDNA microarray which contained 5962 genes (95.5% of the 6241 yeast genes; DNA Chip Re- search Institute). The labeling and hybridization were carried out as de- scribed previously (29). A total of 226 mg of mRNA was used to make fluorescence-labeled cDNA. The labeled cDNA for the exposed (Cy5) and control (Cy3) samples was mixed. After hybridization and washing, the fluorescence image was using a ScanArray 4000 (General Scanning Inc.). The image was analyzed by using Quant Array ver.2.0 (General Scanning Inc.). Microarray data were obtained once for each experimental set (control and exposed). To minimize internal errors, we determined the confidence levels for each differential gene expression as shown below. Microarray Data Mining Many studies have shown that the microarray data will include many false positives and negatives (27), so we estimated the confidence level (CL) of the difference in expression between control and exposed cul- tures. Raw data were normalized by (1) eliminating genes with zeros in either channel (signal or background), (2) keeping only those with a co- efficient of variation (standard deviation/signal) #1.0 for signal, (3) keep- ing only channels with signal above background for all channels, (4) keeping only channels with signal/background .2.0, (5) subtracting background for each channel, (6) normalizing each channel by dividing by the sum, (7) multiplying each signal by 107, and (8) taking ratio and log ratio [5 log2 (ratio)]. After the normalization, 4,400 to 5,500 genes remained in each data set. To determine the CL, we assigned factors to the value for each gene: (1) For signal intensity, we assigned a score of 3 if the intensity was greater than 5,000, a score of 2 for those between 1,000 and 5,000, and 31POWER-FREQUENCY MAGNETIC FIELDS AND GENE EXPRESSION IN YEAST TABLE 1 Continued–Extended Heat shock Protein mRNA ratio CL Minimal Protein mRNA ratio CL MF Protein 10 mT mRNA ratio CL 150 mT mRNA ratio CL 300 mT mRNA ratio CL NI — — NI 2.0 0.7 0.7 1.7 — — — — NI — — NI 1.2 0.9 1.1 ND — — — — NI — — NI 1.3 0.9 0.7 0.7 — — — — 1.1 ND 1.2 1.0 — — — — 0.8 1.1 1.0 1.0 — — — — TABLE 2 Cell Processes and Gene Number Regulated by each Positive Control Condition Aerobic Process Number of genes UP DOWN Heat shock Process Number of genes UP DOWN Minimal medium Process Number of genes UP DOWN cell cycle cell wall biogenesis fatty acid metabolism glycolysis oxidative phosphorylations 3 1 5 6 6 7 4 0 0 1 arginine biosynthesis ATP synthesis cell cycle cell wall biogenesis cytoskeleton 3 2 6 3 2 2 2 1 2 4 cell wall biogenesis chromatin structure cytoskeleton meiosis mitosis 3 3 5 4 4 1 1 1 0 2 protein degradation protein folding protein synthesis purine biosynthesis secretion 3 5 44 7 5 4 0 0 0 1 DNA repair ER and mitochondrial translocation glycolysis methionine biosynthesis or metabolism oxidative phosphorylation 5 4 0 8 9 2 0 12 0 0 mRNA splicing nuclear protein targeting protein degradation protein processing protein synthesis 3 3 5 3 5 1 1 2 1 4 signaling sporulation sterol metabolism transcription transport 2 1 1 6 5 4 4 7 0 7 oxidative stress response protein degradation protein folding protein synthesis respiration 8 3 8 6 2 0 3 1 70 4 secretion trascription transport vacuolar protein targeting 5 7 4 3 2 3 4 2 secretion sterol metabolism stess response TCA cycle transcription transport 4 7 3 0 10 12 3 0 1 5 2 7 others unknown total 37 53 190 61 173 273 75 155 335 389 84 594 40 92 189 28 53 106 Note. This list shows only process including greater than 4 genes with medium, high or highest confidence level. a score of 0.5 for those less than 1000. (2) For the normalized ratio, we assigned a factor of 3 for those greater than 5.0, a score of 2 for those between 3.0 and 5.0, a score of 0.5 for those between 2.0 and 3.0, and a score of 0 for those less than 2.0. (3) For the S/N value of each channel of Cy3 and Cy5, we assigned a score of 3 to those greater than 10.0, a score of 2 for those between 5.0 and 10.0, a score of 1 for those between 2.0 and 5.0, and a score of 0 for those below 2.0. (4) Each score was multiplied to get the final CL. We classified the CLs as follows: highest (36–81), high (18–27), medium (6–16), low (1.5–4.5), lowest (0.25–1.0), or no change (0). PCA and Other Bioinformatic Analysis Principal component analysis (PCA) can be used to explore the vari- ability in gene expression patterns and to identify a small number of themes or principal components. The first principal component is ob- tained by finding the linear combination of expression patterns explaining the greatest amount of variability in the data. The second principal com- ponent is obtained by finding another linear combination of expression patterns that is at right angles to (i.e. orthogonal to and uncorrelated with) the first principal component. The second principal component must ex- plain the greatest amount of the remaining variability in the data after accounting for the first principal component. Each succeeding principal component is obtained similarly. There will never be more principal com- ponents than there are variables (experimental points) in the data. Any individual gene expression pattern can be recreated as a linear combi- nation of the principal component expression patterns. PCA was carried out using GeneSpring (ver. 4.2, Silicon Genetics) and our own program to check the statistical significance of each principal component. To de- termine statistical significance, we use an approach similar to that of Holter et al. (31), who used Singular Value Decomposition (a technique similar to PCA). From the six microarray data sets, six principal components were cal- culated; PC1 accounted for 33.0% of the variance, PC2 accounted for 24.4% of the variance, PC3 accounted for 18.2% of the variance, PC4 accounted for 11.7% of the variance, PC5 accounted for 7.4% of the 34 NAKASONO ET AL. oxidative phosphorylation, protein folding, protein synthe- sis, purine biosynthesis, and transcription-related genes. Some processes were down-regulated, including cell wall biogenesis, fatty acid metabolism, sporulation, and sterol metabolism-related genes. For heat-shock conditions, some cell processes related to adaption to the stress were up- regulated, including cell cycle, DNA repair, ER and mito- chondrial translocation, methionine biosynthesis or metab- olism, oxidative phosphorylation, oxidative stress response, protein folding, and sterol metabolism-related genes. Some process were down-regulated, such as glycolysis, protein synthesis, and TCA cycle-related genes. For minimal me- dium conditions, some cell processes were up-regulated, such as cell wall biogenesis, chromatin structure, cytoskel- eton, meiosis, mitosis, mRNA splicing, nuclear protein tar- geting, protein degradation, and protein processing-related genes. These results suggest that even if most differential gene expression was regulated for unknown reasons, the response of the cell could be found as high-CL differential gene expression. In contrast, for magnetic-field exposure, each distribution of differential gene expressions was very narrow, and no genes were categorized as differentially expressed with high or highest CLs. Only 21, 3 and 3 genes were cate- gorized with medium CLs for 10, 150 and 300 mT, respec- tively. Most of differentially expressed genes seen after magnetic-field exposure were in the low or lowest CLs (Fig. 3) because of the low signal intensity, low S/N ratio, or low expression difference. No up-regulated genes were found with any CL for two of the three magnetic-field in- tensities. No reproducible or dose–response relationship was found. For down-regulated genes, 39 genes were found with CLs in at least two of the three magnetic-field inten- sities. For these genes, only seven showed a dose–response relationship in the normalized ratio data (Table 3). The range of the slopes was from –0.00097 to –0.00059. The range of y intercepts was from 0.47 to 0.69/mT. The range of correlation coefficients was from 0.82 to 0.998. Only three genes showed a reproducible change with CL for all three intensities (Table 3). Gene Expression in Microarray Data The expression of genes related to fundamental cell pro- cesses (heat-shock response, DNA repair, respiration, pro- tein synthesis, cell cycle) is shown in Fig. 4A. In the heat- shock response process, some highest or high CL genes were found for the positive controls. HSP30 and SSA4 were found to be up-regulated in aerobic and minimal medium conditions, respectively. For heat shock, eight genes (SSA1, SSA2, SSA3, SSA4, HSP12, HSP82, HCH1 and SIS1) were found, and the average of the log ratio for 26 genes was higher than for the other conditions. For DNA repair, only one highest-CL gene, OGG1, was found to be down-regu- lated for minimal medium conditions. The average log ratio for heat-shock conditions was slightly higher than for the other conditions. For respiration, only one highest-CL gene was found to be down-regulated for minimal medium con- ditions: The average log ratio for all conditions was close to zero. For protein synthesis, some highest- or high-CL genes were found for positive controls. The average log ratio for heat-shock conditions was lower than for others, and that for aerobic conditions was higher than for others. For the cell cycle, some highest-CL genes were found for aerobic and heat-shock conditions. The average log ratio for heat-shock conditions was a slightly higher than for others. For magnetic-field exposure, only low- or lowest- CL genes, except one gene (PDR13) with medium CL in the heat-shock response, were found. The average log ratio for magnetic-field exposure conditions was almost zero. PCA can be applied to summarize the way in which genes respond collectively under different conditions. Ex- amination of the components can also provide insight into the underlying factors that are measured in the experiment. PCA suggests that much of the variability observed in ex- periments can be summarized by just three components; i.e., three variables capture most of the information in the gene expression pattern. Figure 4B shows three principal components for each condition. The principal components for the positive controls were obviously very different. For magnetic-field exposure, the principal components between the three flux densities were similar, but the difference was statistically significant. There is no dose–response relation- ship between the magnetic-field intensities. Comparison of Changes in Expression between 2D PAGE and Microarray Analysis Table 1 shows a comparison of gene expression data ob- tained by 2D PAGE and microarray analysis. On the 2D images, 62 proteins have been assigned in the SWISS- 2DPAGE database. Fifty-five assigned proteins confirmed microarray data. The remaining seven proteins were not found on the microarrays. Over 40% of the proteins on our 2D gels could not be identified, which may be due to dif- ferences in cell strain or culture conditions. The direction of the changes in expression for some proteins, such as SSA2, PDC1, HSP60, ADE1 (aerobic), SSA1, SSA2, MET6 and HSP60 (heat shock), agreed with the changes in mRNA expression. There was disagreement regarding the changes in expression of the proteins and the mRNAs for SSE1, SSA1, MET6, STI1, RPS31 (aerobic), PDB1, TPI1, ENO2, PDC1, SOD1, TAL1, IPP1 (heat shock), PDC1 and SBA1 (minimal). The disagreement may reflect a failure to identify the protein on the 2D gel, or it might reflect post-transcriptional regulation of protein synthesis. DISCUSSION Many biological mechanisms of interaction of ELF mag- netic-field have been proposed (2). One of the most focused hypotheses is that magnetic fields could be a type of en- vironmental stress (2, 5, 6), much like heat shock (2, 5, 6) 35POWER-FREQUENCY MAGNETIC FIELDS AND GENE EXPRESSION IN YEAST FIG. 4. Gene expression behavior (panel A) in fundamental cell process such as heat shock, DNA repair, respiration, protein synthesis and cell cycle. highest confidence, (; high confidence, 3; medium confidence, m; low confidence, m; lowest confidence, m; no difference, ——; average. Panel B: Comparison of principal components for each condition; (m) PC1; (v) PC2; (m) PC3. or DNA damage (2–4, 14–16). However, no heat-shock- like response was detected in E. coli by a genome-scale method, 2D PAGE, after exposure to a high-intensity ELF magnetic field (up to 14 mT r.m.s., circularly or vertically polarized, 5–100 Hz, for up to 6.5 h) (12). No other known or uncharacterized response was found in these bacteria. The authors noted that their findings did not rule out the possibility of an effect in eukaryotes because there are dif- ferences in the stress responses of bacteria and eukaryotes. In this study, we used yeast as a model eukaryote and the genome-scale screening methods of 2D PAGE (about 1/6 of total expected proteins) and a high-coverage cDNA mi- croarray (that included 95% of total expected genes). We also used higher flux density magnetic fields (up to 300 36 NAKASONO ET AL. mT) inducing electric fields up to 2 V/m of 50 Hz and long- term exposure (for 24 h). These conditions would be ex- pected to cause a reproducible or significant effect because the magnetic-field density was at least several hundred times higher than environmental levels of magnetic fields. This study shows that the high intensity of the magnetic field did not affect the level of either heat-shock proteins or mRNAs, while heat shock produced a significant, repro- ducible effect (Figs. 1, 2 and 4, Table 1). Some authors have reported that magnetic-field exposure increases mutation frequency (3, 4, 14–16), and mecha- nisms suggested to explain that effect include altered rad- ical behavior (15, 29, 30), altered DNA replication (14, 15), and enhanced activity of DNA-reactive agents (30). We re- ported previously that these effects were not reproduced in a bacterial mutation assay with a 50 mT, circularly polar- ized 14-mT magnetic field (31). We also suggested that the possibility of DNA damage in eukaryotes could not be ruled out because of the differences in the mechanisms of DNA repair in bacteria and eukaryotes. In the present study, mutation frequency was not investigated; however, signifi- cant damage to DNA should have resulted in differential expression of the genes involved in DNA repair (20–22). The expression of DNA repair genes did not show any dif- ference for the medium, high or highest CLs. The average log ratios for magnetic-field exposure were near zero (Fig. 4). Positive controls had much more effect on these genes, with some genes expressed at the medium CL. Our results also show that the genes involved in fundamental cell pro- cesses, such as the cell cycle, protein synthesis and respi- ration, were not strongly affected by the exposure to the magnetic field (Fig. 4). Most previous studies of the effects of ELF magnetic fields on gene expression have targeted a limited number of proteins or genes that were related to specific cell pro- cesses (2). When a study revealed no effect, the possibility of the biological effects of the magnetic field on the pro- teins or genes that were not studied still remained. The genome-scale screening methods used in this research can be used to solve this type of problem because of the num- bers of genes or proteins they cover. Changes in the ex- pression of uncharacterized proteins or genes would signal the presence of cell responses which had not yet been de- fined or identified. In the case of yeast, the combination of 2D PAGE (1⁄ expected coverage) and cDNA microarray (95% expected coverage) would be one of the most pow- erful methods for finding an effect on either known and unknown cell responses that would be seen as the differ- ential expression of proteins or mRNAs. In this study, nei- ther known nor unknown cell responses were found after exposure to the high-density magnetic field. No specific cluster in gene expression was found after the magnetic- field exposure. No phenotypic changes such as alterations of the growth and shape of cells were seen after the mag- netic-field exposure (data not shown). Some genes were found to be down-regulated reproduc- ibly after the magnetic-field exposure (Table 2), but the signal intensities were at the noise level. These genes rep- resented only 0.2% of the yeast genes. PCA shows different principal components for different magnetic-field densities, and no dose–response relationship was found. The principal component may be revealing differences in the patterns of the experimental noise in different experimental batches be- cause of the high sensitivity of PCA or the small difference in CL in expression. We cannot exclude the possibility that small differences may occur, because there are some genes that were not covered in this microarray. However, most of the genes, including the stress-response genes, did not change, and some genes that did change had a low CL only at the strong magnetic-field intensity. In future studies, the 5% of the genes that were not tested and the differentially expressed genes with lower CL after the magnetic-field ex- posure should be examined by methods with greater sen- sitivity or by quantitative methods, such as real-time PCR. We conclude that a strong 50 Hz magnetic fields up to 300 mT do not appear to act as general stress factors, like heat shock or DNA damage. The magnetic field did not appear to affect gene expression either for defined cell pro- cesses, such as respiration, cell cycle and protein synthesis, or for unknown cell responses in these model eukaryotic cells. There was no highly significant relationship between magnetic-field intensity and differential gene expression. If magnetic fields induce changes in gene expression in higher organisms, such as humans, either the effect is at a rela- tively low level or it involves a mechanism specific to those organisms, rather than a general response of eukaryocytes. ACKNOWLEDGMENTS We are grateful to Ms. Nao Makino and Ms. Rie Sano for their tech- nical help with 2D PAGE. Received: August 26, 2002; accepted: February 10, 2003 REFERENCES 1. N. Wertheimer and E. Leeper, Electrical wiring configuration and childhood cancer. Am. J. Epidemiol. 109, 273–284 (1979). 2. C. J. Portier and M. S. Wolfe, Eds., Assessment of Health Effects from Exposure to Power-Line Frequency Electric and Magnetic Fields. Publication No. 98-3981, National Institutes of Health, Be- thesda, MD, 1998. 3. J. McCann, F. Dietrich, C. Rafferty and A. O. Martin, A critical review of the genotoxic potential of electric and magnetic fields. Mutat. Res. 297, 61–95 (1993). 4. J. McCann, F. Dietrich and C. Rafferty, The genotoxic potential of electric and magnetic fields: An update. Mutat. Res. 411, 45–86 (1998). 5. O. Smith, Cells, stress and EMFs, Nat. Med. 2, 23–24 (1996). 6. A. Lacy-Hulbert, J. C. Metcalfe and R. Hesketh, Biological responses to electromagnetic fields. FASEB J. 12, 395–420 (1998). 7. P. Cario, B. Greenebaum and E. M. Goodman, Magnetic field ex- posure enhances mRNA expression of s32 in E. coli. J. Cell. Biochem. 68, 1–7 (1998). 8. D. R. Weisbrot, O. Khorkova, H. Lin, A. S. Henderson and R. Good- man, The effect of low frequency electric and magnetic fields on