Nucleic Acids Research

Cell growth conditions

CEFs were prepared from 10-day embryos (C/O) and were maintained in minimal essential medium (MEM) supplemented with 5% (v/v) calf bovine serum, 10% tryptose phosphate broth and 1.0% dimethyl sulfoxide (DMSO) at 38.5°C. NIH3T3 and a packaging cell line, BOSC 23 (10) (ATCC CRL 11270) for ecotropic retrovirus vectors, were maintained in Dulbecco's modified Eagle's medium containing 10% fetal calf serum (DMEM/10% FCS).

Stable transfection

CEFs (1.0 × 10⁶ cells per 60 mm plate) were transfected with 5 µg of the expression plasmids by the polybrene-DMSO method to obtain stable transformants. Two days after transfection, the cultures were transferred and kept under soft agar (0.4%) with 1% DMSO or hard agar (0.7%) with 1.5% DMSO for 7-10 days before counting the number of foci. To concentrate the transformed cell populations in the cultures transfected with pRSV2B-T0.5, -T1 or -T2 DNA, each transfected culture was passaged twice after the detection of foci and kept under soft agar with 1% DMSO. Within a week, almost all the population showed transformed phenotypes, and cells were stocked in liquid nitrogen. Parallel cultures were fixed in 2.0% paraformaldehyde and 0.1% Triton X-100 and immunostained with anti-JunD (C) antiserum. High-level expression of JunD proteins was detected in >95% of the cells as judged from the immunocytochemical analysis.

Plasmid constructions

For the construction of a junD mutant, T0.5, in which the repeat unit (DR2) is duplicated, the 1.1 kb BglII fragment containing the entire wild-type mouse junD coding region preceded by the 5[prime] non-coding sequence from fragment 475 (for higher translation efficiency) was isolated from pRSV2B-junD. To generate pUC119 (B)-junD, this fragment was subcloned into the BglII site of pUC119 (B), which was constructed by introducing a BglII linker at the HincII site of pUC119. The 1.0 kb HindIII (located in the polylinker site)-HincII fragment that contains the 5[prime] half of the junD gene was isolated from pUC119 (B)-junD, and subcloned into the HindIII-HincII site of pHSG398 vector. The resultant plasmid was completely digested at the unique AvaII site (located in DR2; Fig. 1) and ligated with a pair of synthetic oligonucleotides; 5[prime]-GACCTGCACGAGCAGGAGTTCGCCGAAGGCTTCGTCAAGGCGTGGAG-3[prime] and 5[prime]-GTCCTCCAGCGCCTTGACGAAGCCTTCGGCGAACTCCTGCTCGTGCAG-3[prime]. The 1.1 kb HindIII-HincII fragment was isolated from this plasmid and subcloned back into the HindIII-HincII site of pUC119 (B)-junD. The subcloned junD mutant was further introduced into the expression vector, pRSV2B, after BglII digestion to generate pRSV2B-T0.5. The 1.2 kb BglII fragment of pUC 119 (B)-T1 was cloned into the unique BamHI site of pBabe puro (11) to generate pBabe puro T1.

Virus preparation

The replication-competent avian retrovirus vectors carrying the mouse junD and its mutants, T1 and T2, were prepared by DNA transfection into CEFs and subsequent recovery from the culture fluid as described previously (5,9). BOSC 23 cells (2 × 10⁶ cells) on 60 mm dishes were transfected with pBabe puro T1, or pBabe puro using Lipofect Amine reagent according to the manufacturer's protocol, 1 day after seeding. Cells were cultured for 48 h and virus stocks were recovered from the culture fluids. The titers were estimated to be 1-3 × 10⁴ infectious units/ml based on puromycin-resistant colony formation using NIH3T3 as the indicator. The vector stocks were completely free from helper virus as described previously (10).

Molecular cloning and analysis of junD mutants in proviral DNA

Using the PCR technique, the junD sequences were molecularly cloned from chromosomal DNA isolated from the infected CEFs. Two oligonucleotides that cover the 5[prime] and 3[prime] flanking sequences of the single BglII site of pDS3 (9) were synthesized and used as PCR primers. The PCR products were completely digested with BstEII and AccI and separated on agarose gels. The DNA sequences of the BstEII-AccI fragments were determined by the dideoxynucleotide method using a Bcabest sequence kit (TaKaRa).

Protein analysis

The total cellular extracts prepared under denaturing conditions were resolved by electrophoresis on 10% SDS-polyacrylamide gel and JunD protein was detected by western blotting using anti-JunD antibody [anti-JunD (C)] (Santa Cruz Biotechnology) or anti-JunD antiserum [anti-JunD (N)] raised against the synthetic peptide corresponding to amino acids 1-14 of JunD (mouse). Anti-JunD (N) was noncross-reactive to endogenous chicken JunD (12). Bands were visualized by use of the ECL western blotting detection system (Amersham) (13).

RESULTS

The number of repeats in the JunD mutants is changed in the context of a retrovirus vector

To analyze the stability of the junD amplification mutants in chromosomal DNA, T1 (triplication of DR2) and T2 (5-fold multiplication of DR2) were each inserted into a DNA expression plasmid (RSV2B) and transfected into CEFs. The stable transformants of each junD mutant were prepared as mixed populations as described in Materials and Methods and their protein products were analyzed by western blotting using anti-JunD antiserum (Fig. 2A). A single dense band was detected in lysate from each of T1- and T2-transfected CEFs, while no band was detected in nontransfected CEFs. These bands migrated more slowly than the band detected in junD wild type virus-infected CEFs and their apparant molecular weights roughly correspond to those expected, confirming that these mutants are stable in the cellular chromosome.

A    B

Figure 2. (A) Western blot analysis of JunD proteins in virus-infected CEFs or stably transfected CEFs. The cell lysates were resolved by electrophoresis on a 10% SDS-polyacrylamide gel and proteins were detected with anti-JunD(N) antiserum using the ECL western blotting detection system (Amersham). (B) A schematic presentation of the pattern of changes in the DR2 repeat number in the JunD amplification mutants during viral replication. The shaded boxes indicate DR2 units.

We next introduced the entire T1 or T2 coding sequence into plasmids to produce proviral structures of the avian replication-competent retrovirus vector, transfected them into CEFs and recovered stocks of T1 virus and T2 virus, respectively, from the corresponding culture fluids 5 days after transfection. These virus stocks were used to infect fresh CEFs and infected cells were analyzed by western blotting using anti-JunD antiserum. In CEFs infected with these virus stocks, in contrast to the stable transfectants, we detected not only a single protein band of the expected size, but also some additional JunD bands with intermediate mobility between those of JunD wild type and T1 or between those of T1 and T2 (Fig. 2A). Therefore, we next constructed an expression plasmid encoding JunD with duplicated DR2 (designated T0.5; Fig. 1), transfected it into CEFs and isolated the DNA transfectants as described in Materials and Methods. Based on a comparison with the mobility of T0.5, T1 or T2 proteins produced by the DNA-transfected CEFs, T1 virus seemed to produce mainly T1 and T0.5, as well as a small amount of T1.5 (four repeats), while T2 virus seemed to produce mainly T0.5, T1 and T1.5, as well as a small amount of T2. These results suggested that the repeat number of DR2 in the junD coding sequence had changed (mostly decreased) in the retrovirus vectors during their replication (Fig. 2B). To test this possibility, we isolated chromosomal DNA from junD wild-type virus, T1 and T2 virus-infected CEFs, then the mouse junD sequences in the proviral DNA were recovered by PCR and the products were digested with restriction enzymes to discriminate the number of DR2 from the fragment size (Fig. 3). The proviral DNA derived from either T1 or T2 virus-infected CEFs contained a mixture of two major fragments with sizes corresponding to T1 and T0.5 DNA (Fig. 3, lanes 5 and 4, respectively), as well as a faint band corresponding to T1.5 DNA when the film was exposed for a longer period (data not shown). Sequence analysis of the major fragments confirmed that these fragments contained three and two DR2 units, respectively.

Figure 3. Analysis of junD structure in the proviral DNA of T1-, T2- or junD-virus-infected CEFs. DNA fragments covering the entire junD were amplified by PCR, digested completely with both BstEII and AccI and separated on 1.5% agarose gel. PCR templates used were plasmid DNA of pDS3-junD (lane 1), pDS3-T2 (lane 2) or pDS3-T1 (lane 3), or chromosomal DNA of T2 virus-infected CEFs (lane 4), T1 virus-infected CEFs (lane 5) or junD virus-infected CEFs (lane 6). The sizes (bp) of molecular weight markers (lane M) are indicated on the right side. The positions and molecular sizes of the DNA fragments (bp) having the indicated numbers of the repeat unit are shown on the left side.

As described above, the virus stocks used here were collected 5 days after transfection, by which time the viruses initially produced by the transient expression would have propagated into the entire culture, probably through two to four cycles of infection. In view of the procedure used for virus stock preparation, we considered that the genetic changes had occurred during the replication cycle of the retrovirus.

A decrease in the number of DR2 occurs in the process of reverse transcription even in a single cycle of retrovirus replication

For detailed analysis of this genetic instability in the context of the retrovirus genome, we next introduced the entire T1 coding sequence into a plasmid (pBabe puro) carrying the proviral DNA for the production of an MLV-based replication-defective retrovirus vector. Three days after transfection of pBabe puro junD wild type, pBabe puro T1 and pBabe puro (control vector) into BOSC 23, the virus stocks were collected from the culture fluids. Just after virus collection, the transfected BOSC 23 was disrupted under denaturing conditions, and the obtained cell lysates were analyzed by western blotting using anti-JunD antibody (Fig. 4, T). While pBabe puro-transfected BOSC 23 expressed only low levels of human endogenous JunD, which had the same mobility as mouse JunD, transfection of DNA carrying junD wild type or T1 induced a single strong band with the expected mobility.

Figure 4. Expression of JunD mutants in NIH3T3 infected with each replication-defective virus (lane I) or in the transfected BOSC23 (lane T) that was used for the preparation of the corresponding virus stock. BOSC23 was transfected with pBabe puro (control), pBabe puro junD (junD) or pBabe puro T1 (T1) and lysates were prepared immediately after the recovery of the virus stock. NIH3T3 cells were infected twice with these virus stocks at intervals of 8 h, and cell lysates were prepared 3 days after infection. JunD proteins in the cell lysates of NIH3T3 (lane I) or BOSC23 (lane T) were detected by western blotting using anti-JunD(C) antibody.

The recovered ecotropic retrovirus vectors were transduced into NIH3T3, and cell lysates were prepared under denaturing conditions and analyzed by western blotting (Fig. 4, I). Each transduced NIH3T3 expressed similar exogenous JunD bands to those detected in the BOSC 23 that was used for the preparation of the corresponding virus stock. The one clear exception was that a protein band corresponding to T0.5 protein was specifically detected in the NIH3T3 cells transduced with the T1-carrying vector (Fig. 4; TI, compare lanes I and T). By cellular cloning based on resistance to puromycin, a cellular clone expressing T0.5 was isolated among 16 resistant clones transduced with the vector carrying the T1 gene (data not shown). These results clearly show that the loss of one DR2 unit occurred after the transduction of the vectors and not in the process of RNA transcription by DNA-dependent RNA polymerase in the transfected cells from which the virus vectors were produced. Thus, we think that the reduction in the number of DR2 (from three to two) had occurred during the process of reverse transcription or at a point thereafter, such as integration. Reversion to wild-type JunD cannot be accessed by this system because of the presence of endogenous mouse JunD in NIH3T3.

The JunD mutant with duplication of the DR2 unit (T0.5) has weak transforming activity

To examine the transforming activity of T0.5 in detail, CEFs were stably transfected with pRSV2B-T0.5, as well as pRSV2B-junD, -T1 and -T2. While T1 and T2 started to form foci within 1 week after transfection, as reported previously, T0.5 formed foci from 12 days after infection and the focal morphology was much weaker than that of T1 and T2. JunD wild type, however, formed no foci at all, as previously reported (9). At 14 days after transfection, the numbers of foci formed by the transfection of 5 µg DNA of pRSV2B-T0.5, -T1 and -T2 were 2, 17 and 26, respectively (average values of four plates), indicating that duplication of the sequence is sufficient for transforming activity and that the number of DR2 roughly correlates with the strength of the transforming phenotype.

Changes in proviral populations are dependent upon the growth conditions of the infected culture

Since the repeat number of DR2 is readily decreased in each cycle of retrovirus replication, it is of interest to consider why T1 and T2 with three and five DR2 units, respectively, were originally isolated from the replication-competent retrovirus vectors (9). CEF culture harboring the provirus of T1 or T2 was successfully grown from an isolated single focus under soft agar and no virus passage was involved in the original cloning procedure. To examine the roles of selection and unstable mutation in modulating the retrovirus population, we infected the virus stock of T1 into CEFs at a rather low m.o.i. (10^-4) and kept the cells under soft agar to mimic the conditions used for the original isolation of CEF culture harboring T1 and T2. When this soft agar-overlaid culture was incubated for 8 days, strongly transformed cells were accumulated. The agar layer was removed and the cells were transferred to a liquid culture, then virus stocks were collected 1.5 days after the transfer. Western blotting analysis using anti-JunD antiserum revealed that these cultures expressed mainly T1 and T1.5 (Fig. 5, lane 2). A parallel culture that was infected in the same way, but kept in liquid culture instead, expressed mainly T0.5 and T1 (Fig. 5, lane 1), as did CEFs infected with the original T1 virus stock at high m.o.i. (Fig. 5, lane 4). These results indicate that the repeat number of DR2 in junD in the proviral population was increased by the preferential growth of strongly transformed cells in the soft agar-overlaid culture, which appeared to cause biased virus propagation in the culture.

Figure 5. Changes in the proviral population in T1 virus-infected CEFs culture determined by following the expression of JunD proteins. Two cultures of CEFs were infected with the original T1 virus stock at the m.o.i. of 10^-4 and one of them was kept under soft agar for 8 days, then transferred to a liquid culture for 1.5 days for recovery of the virus stock and disrupted just after the virus recovery (lane 2). The other was kept in liquid culture, transferred once and disrupted 1.5 days after the transfer (lane 1). The virus stock prepared as described above and the original T1 virus stock were used to infect fresh CEFs at the m.o.i. of >2, and cell lysates were prepared 2 days after infection (lanes 3 and 4, respectively). JunD proteins were detected as described in the legend to Figure 2A.

When the virus stocks produced by CEFs kept in soft agar (expressing mainly T1 and T1.5) were used to infect fresh CEFs at an m.o.i. of >2, the infected CEFs expressed mainly T0.5 and T1 (Fig. 5, lane 3) and showed significantly weaker transforming morphology compared with the CEFs used for the virus recovery. These results confirmed that the change occurs in a single cycle of replication also in the replication-competent virus, and at a rather higher frequency than we had observed in the replication-defective vectors (Fig. 4; TI, lane I). These results suggest that the rate of reduction in the repeat number of the DR2 unit is affected by the context where the DR2 repeats are located or by enzymatic differences between the reverse transcriptases derived from RSV and MLV. These results also indicate the importance of selective forces that act on proviral populations in culture, allowing the isolation of unstable junD mutants in the retrovirus genome.

DISCUSSION

We have observed rather rapid deletion of the DR2 unit in junD repeat mutants in the context of either the RSV-based replication-competent retrovirus vector (Fig. 2A and Fig. 5) or, although to a lesser extent, of the MLV-based replication-defective retrovirus vector (Fig. 4). DR2 repeats were rather stable in the proviral DNA and neither increase nor decrease in the repeat number was detected in the chicken chromosome, in contrast to the case of GC-rich triplet repeats in the human genome, where a high frequency of length alteration is observed (1,2).

The rate of genetic change observed here was much higher than that of classical point mutations detected in the retrovirus genome; such genetic instability has never been observed in c-src (14) point mutants that have acquired transforming activity during propagation of the same avian retrovirus vector. Therefore, marked instability seems to be a unique feature of repeat mutants. Deletion of the DR2 units was shown to occur during the process of reverse transcription (15,16), though we cannot fully exclude the possibility that some other process such as integration is also involved. Two models, which are not mutually exclusive, can be considered to explain the molecular mechanisms of these genetic changes generated by reverse transcriptase (Fig. 6A and B). In the tandem repeat of DR2, the nascent strand growing point of reverse transcription may switch template to the other template present in the same virion (Fig. 6A), or to another identical sequence in the same template (Fig. 6B). In both models, the template could be either (+) RNA or (-) DNA. Since DR2 has a high GC content (65%) and contains two pairs of inverted repeat sequences composed of 4 and 5 bp (Fig. 1), secondary structures which are favorable for looping out would be formed in DR2 in (+) RNA (Fig. 6C) or (-) DNA, and should enhance the frequency of the jump to the same template (17) (Fig. 6B). In this respect, it is interesting that hairpin structures have been suggested to play important roles in triplet repeat expansion events in DNA replication (2).

Figure 6. Possible mechanisms of change in the number of tandem repeats during the process of reverse transcription. The retrovirus genome (+) RNA is shown by solid lines and (-) DNA is shown by dotted lines. Each arrow indicates one unit of the repeat. The jumping of the reverse transcription point occurs (A) to the other template in the nucleocapsid or (B) within the same template. These models are equally applicable to (+) DNA synthesis from the (-) DNA template. (C) The schematic presentation of one of the possible secondary structures formed in the tandem repeat sequences of (+)RNA.

In order for an increase in DR2 repeat number to occur in the latter model (Fig. 6B), the detachment of the entire DR2 from the base-paired template is required at the growing point of the nascent strand (17). Therefore, this model would favor a decrease of the tandem repeat number, since this case does not involve a large disturbance of the hydrogen bonding (17). On the other hand, the former model could produce either an increase or a decrease (16,18). Considering the fact that proviral DNAs carrying either T1 or T2 were isolated from the same transformed CEF clone, it is reasonable to speculate that multiplication (increase in the number of repeats) could have occurred before the isolation of T2, probably through the mechanism shown in Figure 6A. Indeed, we have observed generation of T1.5 in the propagation step of the T1 virus stock (Fig. 2A). The frequency of multiplication, however, seems to be much lower than that of deletion.

The strong transforming phenotypes assumed by CEFs expressing these repeat mutants were not successfully transferred by infecting fresh CEFs with virus stocks obtained from these culture fluids. This result can be well explained by our findings that the repeat number of the junD mutants can be easily reduced within a single cycle of viral replication (Figs 4 and 5) and that the repeat number (from two to five) roughly correlates with the strength of the transforming activity. These findings are also consistent with the difficulty that we encountered in our original trial for the virological cloning of these junD mutants; strongly transforming phenotypes were not efficiently transferred by viral infection. We therefore used PCR cloning directly from the chromosomal DNA isolated from the strongly transforming CEFs (9).

Other examples of duplication (or less frequently, triplication) mutants in retroviruses are known (19-23). They include duplication of nine oligonucleotides in the coding sequence of the v-ros oncogene carried by the avian sarcoma virus UR2 (20), and duplication of oligonucleotides in the long terminal repeat of an erythroleukemic retrovirus (SFFV) (22) and mink cell focus-forming (MCF) virus (23). It is worth noting that all these mutants are expected to have a growth advantage or leukemogenic potential for the target cells, and that at least some of them are known to be unstable (22). Therefore, it is quite possible that some multiplication mutants can be detected as a consequence of their strong growth advantage and pathogenic effects, in spite of their instability. For the detection of pathologically important amplification mutations, therefore, direct analysis of the proviral structure in the target tissues would be much more reliable than cloning of the virus sequence from the virion present in the patient.

ACKNOWLEDGEMENTS

We thank Drs T. Kitamura and T. Heike for their kind advice on replication-defective viruses. We are grateful for Drs M. Yamada and N. Gotoh for fruitful discussions and for providing NIH3T3 cells, respectively. We thank Etsuko Endo and Yuriko Yoshikawa for assistance in the preparation of the manuscript. This work was supported in part by Grants and Endowments from Eisai Co., Ltd and by a Grant-in-Aid for Scientific Research on Priority Areas from the Ministry of Education, Science and Culture, Japan.

REFERENCES

1. The Huntington's Disease Collaborative Research Group (1993) Cell, 72, 971-983.

2. Mitas,M., Yu,A., Dill,J., Kamp,J., Chambers,E.J. and Haworth I.S. (1995) Nucleic Acids Res., 23, 1050-1059. MEDLINE Abstract

3. Hartl,M. and Vogt,P.K. (1992) Cell Growth Differ., 3, 909-918. MEDLINE Abstract

4. Maki,Y., Bos,T.J., Davis,C., Starbuck,M. and Vogt,P.K. (1987)Proc. Natl Acad. Sci. USA, 84, 2848-2852. MEDLINE Abstract

5. Okuno,H., Suzuki,T., Hashimoto,Y., Curran,T. and Iba,H. (1991) Oncogene, 6, 1491-1497. MEDLINE Abstract

6. Suzuki,T., Hashimoto,Y., Okuno,H., Sato,H., Nishina,H. and Iba,H. (1991) Jap J. Cancer Res., 82, 58-94.

7. Abate,C., Luk,D., Gange,E., Roeder,R.G. and Curran,T. (1990)Mol. Cell Biol., 11, 3624-3632.

8. Nishimura,T. and Vogt,P.K. (1988) Oncogene, 3, 659-663. MEDLINE Abstract

9. Kameda,T., Akahori,A., Sonobe,M.H., Suzuki,T., Endo,T. and Iba,H. (1993) Proc. Natl Acad. Sci. USA, 90, 9369-9373. MEDLINE Abstract

10. Pear,W.S., Nolan,G.P., Scott,M.L. and Baltimore,D. (1993)Proc. Natl Acad. Sci. USA, 90, 8392-8396. MEDLINE Abstract

11. Morgenstern,J.P. and Land,H. (1990) Nucleic Acids Res., 18, 3587-3596. MEDLINE Abstract

12. Hartl,M., Hutchins,J.T. and Vogt,P.K. (1991) Oncogene, 6, 1622-1631.

13. Murakami,M., Sonobe,M.H., Ui,M., Kabuyama,Y., Watanabe,H., Wada,T., Handa,H. and Iba,H. (1997) Oncogene, 14, 2435-2444. MEDLINE Abstract

14. Iba,H., Takeya,T., Cross,F.R., Hanafusa,T. and Hanafusa,H. (1984)Proc. Natl Acad. Sci. USA, 81, 4424-4428. MEDLINE Abstract

15. Coffin,J.M. (1992) Curr. Top. Microbiol. Immunol., 176, 143-164. MEDLINE Abstract

16. Stuhlmann,H. and Berg,P. (1992) J. Virol., 66, 2378-2388. MEDLINE Abstract

17. Pathak,V.K. and Temin,H.M. (1990) Proc. Natl Acad. Sci. USA, 87, 6024-6028. MEDLINE Abstract

18. Pathak,V.K. and Temin,H.M. (1990) Proc. Natl Acad. Sci. USA, 87, 6019-6023. MEDLINE Abstract

19. Myers,G., Rabson,A.B., Josephs,S.F., Smith,T.F., Berzofsky,J.A. and Wong-Staal,F. (1990) Human Retroviruses and AIDS. Los Alamos National Laboratory, Los Alamos, NM.

20. Neckammeyer,W.S., Shibuya,M., Hsu,M.-T. and Wang,L.-H. (1986)Mol. Cell. Biol., 6, 1478-1486.

21. Rhode,B.W., Emerman,M. and Temin,H.M. (1987) J. Virol., 61, 925-927. MEDLINE Abstract

22. Spiro,C., Li,J., Bestwick,R.K. and Kabat,D. (1988) Virology, 164, 350-361. MEDLINE Abstract

23. Stoye,J.P., Moroni,C. and Coffin,J.M. (1991) J. Virol., 65, 1273-1285. MEDLINE Abstract

---

*To whom correspondence should be addressed. Tel: +81 3 5449 5730; Fax: +81 3 5449 5449; Email: iba@ims.u-tokyo.ac.jp

---

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 17 Oct 1998
Copyright©Oxford University Press, 1998.
CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				T. Ito, M. Yamauchi, M. Nishina, N. Yamamichi, T. Mizutani, M. Ui, M. Murakami, and H. Iba Identification of SWI{middle dot}SNF Complex Subunit BAF60a as a Determinant of the Transactivation Potential of Fos/Jun Dimers J. Biol. Chem., January 19, 2001; 276(4): 2852 - 2857. [Abstract] [Full Text] [PDF]

This Article Abstract Print PDF (108K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (2) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Ito, T. Articles by Iba, H. Search for Related Content PubMed PubMed Citation Articles by Ito, T. Articles by Iba, H. Social Bookmarking          What's this?