Skip Navigation


Bioinformatics Advance Access originally published online on November 30, 2007
Bioinformatics 2008 24(3):303-308; doi:10.1093/bioinformatics/btm589
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (Print PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
24/3/303    most recent
btm589v1
Right arrow Comments: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when Comments are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Fujita, S.
Right arrow Articles by Iba, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fujita, S.
Right arrow Articles by Iba, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Putative promoter regions of miRNA genes involved in evolutionarily conserved regulatory systems among vertebrates

Shuji Fujita and Hideo Iba *

Division of Host-Parasite Interaction, Department of Microbiology and Immunology, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minato-ku Tokyo, 108-8639, Japan

*To whom correspondence should be addressed.


   ABSTRACT
TOP
 ABSTRACT
1 INTRODUCTION
 2 METHODS
 3 RESULTS AND DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 

Motivation: Just as transcription factors, miRNA genes modulate global patterns of gene expression during differentiation, metabolic activation, stimulus response and also carcinogenesis. However, little is currently known how the miRNA gene expression itself is regulated owing to lack of basic information of their gene structure. Global prediction of promoter regions of miRNA genes would allow us to explore the mechanisms underlying gene-regulatory mechanisms involving these miRNAs.

Results: We speculate that if specific miRNA molecules are involved in evolutionarily conserved regulatory systems in vertebrates, this would entail a high level of conservation of the promoter of miRNA gene as well as the miRNA molecule. By our current screening of putative promoter regions of miRNA genes (miPPRs) on this base, we identified 59 miPPRs that would direct production of 79 miRNAs. We present both biochemical and bioinformatical verifications of these putative promoters.

Contact: iba{at}ims.u-tokyo.ac.jp

Supplementary information: Supplementary data are available at Bioinformatics online.


   1 INTRODUCTION
TOP
 ABSTRACT
 1 INTRODUCTION
2 METHODS
 3 RESULTS AND DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
MicroRNAs (miRNAs) are endogenous 20–24 nt RNAs known to mediate the repression of target mRNAs by suppressing translation or promoting mRNA decay in animal (Zamore and Haley, 2005). More than 500 species of miRNAs have now been identified in human (Griffiths-Jones et al., 2006), and also predicted to target roughly 30% of the total coding genes in human (Lewis et al., 2005; Xie et al., 2005). In spite of the strong impact of miRNAs on regulatory network comparable to transcription factors, it still remains largely unknown how human miRNA expression itself is regulated at the transcriptional level, although the vertebrate miRNA genes are thought to be generally transcribed by RNA polymerase II (pol II) to produce a pri-miRNA containing a 5'-cap structure and polyA tail (Cai et al., 2004; Lee et al., 2004). Several reports have now shown examples of vertebrate miRNA gene regulation by known transcription factors (O’Donnell et al., 2005; Zhao et al., 2005). Some overrepresented motifs have been also described in limited upstream flanking regions of miRNA embedding regions in nematodes, plants and mammals (Lee et al., 2007; Ohler et al., 2004; Zhou et al., 2007). However in these studies, miRNA gene structures and transcription units were not considered. It would be partly because identification of a transcription unit of miRNA is very difficult owing to low expression levels of certain miRNA genes in specific tissues (Farh et al., 2005), and to instability of primary transcripts of miRNAs (pri-miRNAs) (Cullen, 2004). Therefore, currently only a few pri-miRNA structures have been described biochemically (Cai et al., 2004; Kim and Kim, 2007; Lee et al., 2004; Taganov et al., 2006). Some miRNAs embedding regions are reported to reside in introns of certain coding genes (Kim and Kim, 2007; Rodriguez et al., 2004), but it is not fully understood how these miRNA are produced.

To understand regulatory networks that involve miRNAs, we are interested in important regulatory systems conserved across vertebrates. We speculated that in such miRNA molecules, a high level of conservation would be required for not only the miRNA molecule itself but for the promoter of the miRNA gene. We have performed a computational approach to predict promoters of such miRNAs on this basis by use of comparative genomic methods that have been successfully employed to some extent in the elucidation of functional cis-regulatory elements of coding genes (Blanchette and Tompa, 2002; Loots et al., 2002; Xie et al., 2005). We finally present a validity of our prediction using bioinformatical and biochemical analyses.


   2 METHODS
TOP
 ABSTRACT
 1 INTRODUCTION
 2 METHODS
3 RESULTS AND DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
The databases used in this study are described in online Supplementary Methods. We first searched for aligned blocks containing miRNA hairpin regions that were conserved between either human and chicken or between human and zebrafish in the Blastz-alignment at UCSC (Blanchette et al., 2004). We set a threshold of the original alignment scores at higher than 50 000 to filter out non-conserved miRNAs and to remove overly short or poorly conserved blocks. When a hairpin region was split into two adjacent blocks, we concatenated the blocks.

We next searched for highly conserved blocks within the 100 kb upstream of the conserved blocks containing miRNA genes, with the same degree of conservation as above. Assuming that a hairpin region of miRNA is embedded in the middle of the miRNA gene, a limitation of 100 kb is reasonable as 95% of the coding genes in RefGene track are shorter than 200 kb. We removed the blocks whose chromosome are different from the ones containing miRNA to assure that each block is located upstream of miRNA in each species. Using FASTA (Pearson and Lipman, 1988), we also removed the upstream blocks in cases where the corresponding human sequence shares homology with the CDS or 3'UTR of human mRNA sequences (RefSeq) (≥85% sequence identity).

From the remaining upstream blocks, we further removed blocks containing other conserved miRNA hairpin regions. When the adjacent consensus regions were ≤300 bp apart, the blocks were concatenated. Consensus sequences were then extracted from the selected blocks as follows: in the position where the base is conserved either between human and chicken or between human and zebrafish and in addition between human and at least one other species among mouse, cow, lizard, chicken, frog, medaka and zebrafish, the base is regarded as a consensus base. Non-conserved bases or gaps are represented by ‘n’.

We searched for core promoter elements in the consensus region by an entropy-based calculation (Frech et al., 1993; Schneider et al., 1986) with TRANSFAC weight matrices (TATA box: M00216, M00252, M00471, CAAT box: M00254 and GC box: M00255) according to the following equations;


Formula 1

(1)


Formula 2

(2)


Formula 3

(3)
where Ci(i), f(i,b), m and Sk denote the consensus index at position i, relative frequency of base b at position i, and length of matrix and similarity of sequence at kth slid window to consensus one, respectively.

The following cut-off values of similarity were used: Sk > 88.6 for TATA box, Sk > 87.6 for CCAAT box and Sk > 86.2 for GC box. Details regarding our determination of the threshold are described in the Supplementary Materials and Methods. After selecting blocks containing at least one core promoter element, we finally selected the block that was located in the closest proximity to the corresponding miRNA and defined it as an miPPR. All computations were run on 96-CPUs Sun Fire 15K systems and a 128-CPUs Linux cluster system.


   3 RESULTS AND DISCUSSION
TOP
 ABSTRACT
 1 INTRODUCTION
 2 METHODS
 3 RESULTS AND DISCUSSION
ACKNOWLEDGEMENTS
 REFERENCES
 
We made two critical assumptions during the screening for promoters of miRNAs operating in evolutionary conserved systems across vertebrates. First, such miRNAs would not only be derived from pre-miRNA sequences that are conserved among vertebrates, but would also have a conserved promoter sequence. Second, such a conserved promoter harbours at least one core promoter element for RNA polymerase II (TATA, CCAAT and GC box). We screened such conserved regions within a 100 kb region upstream of a series of miRNAs that are conserved among vertebrates, and defined miPPR as one which is in closest proximity to the corresponding miRNA embedding region (see Supplementary Fig. S1). ‘Conservation among vertebrates’ in this method indicates that conservation is detectable at least between either between human and chicken or between human and zebrafish. Details of the algorithm are described in Methods.

By this screening, 59 regions were eventually selected as the miPPRs for 79 miRNAs. These miPPRs are summarized in Table 1 together with their genomic coordinates, their length (bp) as well as the closest distance between miPPR and the corresponding miRNA hairpin region. The consensus sequences of miPPRs are available in Supplementary Material. Some of these different miRNAs share the same miPPR, suggesting that they are generated as polycistronic transcripts. Among them, such miR-1-2/-133a-1 cluster (Rao et al., 2006; Sempere et al., 2004) or miR-106a/-18b/-20b/-19b-2 clusters (Fontana et al., 2007), for example, has been shown to be co-regulated in several tissues.


View this table:
[in this window]
[in a new window]

  		Table 1. Listing of the miPPRs identified in this study, each of which was named after the most proximal miRNA

 
The median distance between the human miPPR position and the corresponding miRNA was estimated to be 5.5 kb. By assuming that the hairpin region of the miRNA is located in the middle of the miRNA gene, this distance is less than one half of the median length of the known coding genes (11.4 kb). As shown in Figure 1A, the identified human miPPRs positions relative to the corresponding miRNAs are significantly shorter when compared with the 20 sets of background observation, where background PPRs were found from randomly selected conserved blocks as cohorts against miRNAs using the same procedure as miPPRs search. The distances between 22 miPPRs and their corresponding miRNA hairpin regions are longer than 10 kb, whereas 5 miPPRs is present very close to them (≤25 bp) (Supplementary Table S1). We believe that these miPPRs are also good candidates for the miRNA promoter and discussed the basis for it partly using computational analysis in Supplementary Note. Moreover, calculation of odds ratio of CpG dinucleotides ({rho}CpG) (Nakashima et al., 1997) shows that the human miPPRs had 1.43- and 1.64-fold higher {rho}CpG than that of the background cohorts described above and that of entire human genome, respectively (Fig. 1B). Considering that a significant fraction of known promoters locate in CpG island, these results support that the miPPRs are rich in functional promoter sequences.


Figure 1
View larger version (49K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. The statistical (A and B) and experimental (C–J) analyses on human miPPRs. (A) Difference between distance distribution of miPPRs from the corresponding miRNA and that of 20 background sets [one-tailed KS-test; median (IQR) is 1.9 x 10–5 (5.5 x 10–6–5.1 x 10–5)]. (B) The odds ratios of CpG dinucleotides (left) and the percentage of GC content (right) in the human sequences of miPPRs, the human sequences of the conserved regions of 20 background sets (the same as background sets in A), the randomly selected human genomic sequences (20 sets) and the known human promoter sequences. (C–D) Two examples of miPPRs that correspond to the miRNA promoters reported previously. Genomic loci of miPPR-146a (C) and miPPR-126 (D) are shown together with the EST densities of human (EST) and the degree of conservation (Cons). (E–J) Structural analysis and genomic loci of miPPR-1-2 (E–G), miPPR-199a-2 (H and I) and miPPR-21 (J), respectively. (E) Determination of transcription start site (TSS) by primer extension. In left, products of dideoxynucleotide sequencing of antisense strand using the same primers were comigrated to determine precise TSS. The sequence of template strand is shown with TATA box and the determined TSS. (F, H) Northern blot analysis. The used probes are shown in G or in I, respectively. Genomic loci of miPPR-1-2 (G), miPPR199a-2 (I) and miPPR-21 (J) are shown together with determined structures, respectively.

 
Although only a few miRNA promoters has been reported, we found that the reported promoter regions for miR-146a (Taganov et al., 2006), miR-126 (Kim and Kim, 2007) and miR-10b (Ma et al., 2007) are consistent with the miPPR-146a (Fig. 1C), miPPR-126 (Fig. 1D) and miPPR-10b predicted here, respectively. MiPPR-146a contained the conserved NF-{kappa}B binding sites that were reported to be critical for LPS response (Supplementary Fig. S2A). In the case of miR-10b, TWIST1, a metastasis-promoting transcription factor, has been shown to induce miR-10b via binding to the most proximal E-box upstream of the miR-10b hairpin region. Importantly, we detect this E-box in human miPPR-10b (Supplementary Fig. S3).

We next performed direct biochemical analysis of miPPR-1-2 (for miR-1-2 and miR-133a-1), miPPR-199a-2 (for miR-199a-2) and miPPR-21 (for miR-21) based upon unique expression patterns of these miRNAs, which would be conducive for experimental analysis (Fig. 1E–J). The expression patterns of miR-1 and miR-133a are well characterized in muscle development (Rao et al., 2006; Zhao et al., 2005). Both up- and down-regulation of miR-199a has been reported in several cancer cell types, indicating its potential involvement in tumorigenesis (Volinia et al., 2006). MiR-21 is also highly expressed in various human cancers (Volinia et al., 2006).

Primer extension experiment using human muscle RNA shows a transcription 20 bp downstream of a conserved TATA box in miPPR-1-2 (Fig. 1E). By northern analysis, a single transcript of 5 kb was detected by probes for the just downstream region of miPPR-1-2 and also for the region embedding either miR-1-2 or miR-133a-1, but not by a probe for the just upstream region of miPPR-1-2 (Fig. 1F and G). These results strongly support our contention that miPPR-1-2 indeed functions as the common promoter for miR-1-2 and miR-133a-1. In addition, this showed direct evidence to support the recent suggestion that miR-1-2 and miPPR-133a-1 might be embedded in a polycistronic transcript as described above. In the consensus sequence of miPPR-1-2 (Supplementary Fig. S2B), we also identified the previously reported SRF and MyoD element (Zhao et al., 2005). Similar northern analysis using RNA from HeLa cells has suggested that miPPR-199a-2 also directs a transcript embedding miR-199a-2 (Figs 1H and I). By primer extension and northern analysis, we elucidated that miPPR-21 drives pri-miR-21 30 bp downstream of TATA box present in miPPR-21 (Fig. 1J). These analyses on miR-21 indicated that the authentic pri-miR-21 is transcribed just downstream of TATA box in miPPR-21 but is not drived by the previously reported region (Cai et al., 2004) and the details of biochemical analyses will be published elsewhere with promoter analysis including transcription factors involved in the regulation.

We categorized miPPRs with validation status as a measure of predictive accuracy (Table 1). Our direct biochemical analyses of a primary transcript of miR-1-2 and miR-133a-1 (Figs 1E–G), miR-199a-2 (Figs 1H–I) or miR-21 (Fig. 1J) were consistent with our predicted miPPRs (status 1). miPPRs-146a, miPPR-126 and miPPR-10b are consistent with previous studies of the promoter regions of the miR-146a, miR-126 and miR-10b, respectively (Figs 1C and D) (status 2).

Other than these 6 miPPR, we identified 30 miPPRs that direct transcriptional initiation according to the UCSC database (status 3, 3' and 4) among total 59 miPPRs. Moreover, 19 out of these 30 miPPRs direct transcripts that harbour the corresponding miRNA (status 3) or transcripts that are produced by splicing of introns containing the miRNA (status 3'). All these findings support the validity of our approach. In some miPPRs with no varidation status, we found transcripts containing the miRNAs that are initialized either far upstream or downstream of the miPPR in the database. Therefore, some other promoter candidates would be major promoters, at least in the tissue from which the EST was isolated.

Since 0.03 quantile of the distribution of distances between adjacent coding genes is calculated to be 1.5 kbp, we tentatively regarded that a certain miRNA gene is overlapped with a coding gene in the genome when we find a coding gene locus within 1.5 kb flanking regions of the miRNA embedding region. As shown in Table 1, 30 conserved miRNAs are mapped to be overlapped with coding genes. Among them, miPPR-490 corresponds to the common promoter of the miR-490 gene and the overlapping coding gene, CHRM2, whereas the miR-126 and EGFL7 genes share the common promoter, miPPR-126. These results suggest that the coding gene and miRNA could be co-regulated like these cases. However, most of these miPPRs were located distinct positions from the promoters of the overlapping gene, indicating that expression of miRNA would be regulated independent of the overlapping coding genes in these cases. For example, miR-21 were overlapped with the TMEM49 coding gene, however the expression of miR-21 is regulated by its own promoter, miPPR-21 that resides in an intron of TMEM49, independent of the transcription of TMEM49 (Fig. 1J).

Deciphering transcriptional regulation on miRNA genes is now urgent to understand regulatory networks structured with several layers including miRNA-mediated post-transcriptional repression. Biochemical promoter analysis using this information would be quite efficient and the accumulation of such biochemical data would be useful to improve the algorithm. We believe that such processes will contribute to find out conserved regulatory systems operating in vertebrates for further understanding of miRNA expression potential and misregulation in cancer or other diseases.


   ACKNOWLEDGEMENTS
TOP
 ABSTRACT
 1 INTRODUCTION
 2 METHODS
 3 RESULTS AND DISCUSSION
 ACKNOWLEDGEMENTS
REFERENCES
 
Computation time was provided by the Super Computer System, Human Genome Center, Institute of Medical Science, University of Tokyo. This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas and by the special coordination Funds for Promoting Science and Technology from the Ministry of Education, Culture, Sports, Science and Technology Japan.

Conflict of Interest: none declared.


   FOOTNOTES
 
Associate Editor: Limsoon Wong

Received on October 2, 2007; revised on November 7, 2007; accepted on November 23, 2007

   REFERENCES
TOP
 ABSTRACT
 1 INTRODUCTION
 2 METHODS
 3 RESULTS AND DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 

    Blanchette M, et al. Aligning multiple genomic sequences with the threaded blockset aligner. Genome Res (2004) 14:708–715.[Abstract/Free Full Text]

    Blanchette M, Tompa M. Discovery of regulatory elements by a computational method for phylogenetic footprinting. Genome Res (2002) 12:739–748.[Abstract/Free Full Text]

    Cai X, et al. Human microRNAs are processed from capped, polyadenylated transcripts that can also function as mRNAs. RNA (2004) 10:1957–1966.[Abstract/Free Full Text]

    Cullen BR. Transcription and processing of human microRNA precursors. Mol. Cell (2004) 16:861–865.[CrossRef][Web of Science][Medline]

    Farh KK, et al. The widespread impact of mammalian MicroRNAs on mRNA repression and evolution. Science (2005) 310:1817–1821.[Abstract/Free Full Text]

    Fontana L, et al. MicroRNAs 17-5p-20a-106a control monocytopoiesis through AML1 targeting and M-CSF receptor upregulation. Nat. Cell Biol (2007) 9:775–787.[CrossRef][Web of Science][Medline]

    Frech K, et al. Computer-assisted prediction, classification, and delimitation of protein binding sites in nucleic acids. Nucleic Acids Res (1993) 21:1655–1664.[Abstract/Free Full Text]

    Griffiths-Jones S, et al. miRBase: microRNA sequences, targets and gene nomenclature. Nucleic Acids Res (2006) 34:D140–D144.[Abstract/Free Full Text]

    Kim YK, Kim VN. Processing of intronic microRNAs. EMBO J (2007) 26:775–783.[CrossRef][Web of Science][Medline]

    Lee J, et al. Regulatory circuit of human microRNA biogenesis. PLoS Comput. Biol (2007) 3:e67.[CrossRef][Medline]

    Lee Y, et al. MicroRNA genes are transcribed by RNA polymerase II. EMBO J (2004) 23:4051–4060.[CrossRef][Web of Science][Medline]

    Lewis BP, et al. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell (2005) 120:15–20.[CrossRef][Web of Science][Medline]

    Loots GG, et al. rVista for comparative sequence-based discovery of functional transcription factor binding sites. Genome Res (2002) 12:832–839.[Abstract/Free Full Text]

    Ma L, et al. Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature (2007) 449:682–688.[CrossRef][Medline]

    Nakashima H, et al. Differences in dinucleotide frequencies of human, yeast, and Escherichia coli genes. DNA Res (1997) 4:185–192.[Abstract]

    O’Donnell KA, et al. c-Myc-regulated microRNAs modulate E2F1 expression. Nature (2005) 435:839–843.[CrossRef][Medline]

    Ohler U, et al. Patterns of flanking sequence conservation and a characteristic upstream motif for microRNA gene identification. RNA (2004) 10:1309–1322.[Abstract/Free Full Text]

    Pearson WR, Lipman DJ. Improved tools for biological sequence comparison. Proc. Natl Acad. Sci. USA (1988) 85:2444–2448.[Abstract/Free Full Text]

    Rao PK, et al. Myogenic factors that regulate expression of muscle-specific microRNAs. Proc. Natl. Acad. Sci. USA (2006) 103:8721–8726.[Abstract/Free Full Text]

    Rodriguez A, et al. Identification of mammalian microRNA host genes and transcription units. Genome Res (2004) 14:1902–1910.[Abstract/Free Full Text]

    Schneider TD, et al. Information content of binding sites on nucleotide sequences. J. Mol. Biol (1986) 188:415–431.[CrossRef][Web of Science][Medline]

    Sempere LF, et al. Expression profiling of mammalian microRNAs uncovers a subset of brain-expressed microRNAs with possible roles in murine and human neuronal differentiation. Genome Biol (2004) 5:R13.[CrossRef][Medline]

    Taganov KD, et al. NF-kappaB-dependent induction of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proc. Natl Acad. Sci. USA (2006) 103:12481–12486.[Abstract/Free Full Text]

    Volinia S, et al. A microRNA expression signature of human solid tumors defines cancer gene targets. Proc. Natl Acad. Sci. USA (2006) 103:2257–2261.[Abstract/Free Full Text]

    Xie X, et al. Systematic discovery of regulatory motifs in human promoters and 3' UTRs by comparison of several mammals. Nature (2005) 434:338–345.[CrossRef][Medline]

    Zamore PD, Haley B. Ribo-gnome: the big world of small RNAs. Science (2005) 309:1519–1524.[Abstract/Free Full Text]

    Zhao Y, et al. Serum response factor regulates a muscle-specific microRNA that targets Hand2 during cardiogenesis. Nature (2005) 436:214–220.[CrossRef][Medline]

    Zhou X, et al. Characterization and identification of microRNA core promoters in four model species. PLoS Comput. Biol (2007) 3:e37.[CrossRef][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Clin. Cancer Res.Home page
N. Yamamichi, R. Shimomura, K.-i. Inada, K. Sakurai, T. Haraguchi, Y. Ozaki, S. Fujita, T. Mizutani, C. Furukawa, M. Fujishiro, et al.
Locked Nucleic Acid In situ Hybridization Analysis of miR-21 Expression during Colorectal Cancer Development
Clin. Cancer Res., June 15, 2009; 15(12): 4009 - 4016.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
T. Haraguchi, Y. Ozaki, and H. Iba
Vectors expressing efficient RNA decoys achieve the long-term suppression of specific microRNA activity in mammalian cells
Nucleic Acids Res., April 1, 2009; 37(6): e43 - e43.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (Print PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
24/3/303    most recent
btm589v1
Right arrow Comments: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when Comments are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Fujita, S.
Right arrow Articles by Iba, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fujita, S.
Right arrow Articles by Iba, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?