• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Jan 31, 2006; 103(5): 1313–1318.
Published online Jan 23, 2006. doi:  10.1073/pnas.0508658103
PMCID: PMC1345707
Developmental Biology

Real-time imaging of the somite segmentation clock: Revelation of unstable oscillators in the individual presomitic mesoderm cells


Notch signaling components such as the basic helix-loop-helix gene Hes1 are cyclically expressed by negative feedback in the presomitic mesoderm (PSM) and constitute the somite segmentation clock. Because Hes1 oscillation occurs in many cell types, this clock may regulate the timing in many biological systems. Although the Hes1 oscillator is stable in the PSM, it damps rapidly in other cells, suggesting that the oscillators in the former and the latter could be intrinsically different. Here, we have established the real-time bioluminescence imaging system of Hes1 expression and found that, although Hes1 oscillation is robust and stable in the PSM, it is unstable in the individual dissociated PSM cells, as in fibroblasts. Thus, the Hes1 oscillators in the individual PSM cells and fibroblasts are intrinsically similar, and these results, together with mathematical simulation, suggest that cell-cell communication is essential not only for synchronization but also for stabilization of cellular oscillators.

Keywords: Hes1, oscillation, luciferase, basic helix-loop-helix gene

Somites, precursors for the segmental structures such as the vertebral column, ribs, and skeletal muscles, are generated in a head-to-tail order by periodic segmentation of the anterior end of the presomitic mesoderm (PSM). This periodic event is regulated by the somite segmentation clock, which is composed of Notch and Wnt signaling molecules (1-6). In the PSM, Notch components such as the basic helix-loop-helix genes Hes1 and Hes7 are cyclically expressed, and each cycle leads to segmentation of a bilateral pair of somites (7-11). This oscillatory expression occurs in a synchronous manner but with the caudal-to-rostral phase delay, resulting in wave-like propagation of the expression domains from the caudal to the rostral direction. It has been shown that this oscillatory expression depends upon a negative feedback loop (12-18).

Interestingly, Hes1 oscillation occurs in many cell types in addition to the PSM after serum treatment or activation of Notch signaling, suggesting that this clock may regulate the timing in many biological systems (12). Although Hes1 oscillation is stable in the PSM, it is damped after three to six cycles in other cells, raising the possibility that the Hes1 oscillator of the PSM cells is intrinsically different from that of other cell types (8, 12). However, the damping could result not only from damped oscillation in each cell but also from desynchronization between cells, and it is not clear which is the case. It was shown that PSM cells could become desynchronized when they are dissociated (19), but the nature of the segmentation clock in individual PSM cells remains to be determined.

To understand the dynamics of the somite segmentation clock, we attempted real-time imaging of Hes1 expression in the PSM and the dissociated individual PSM cells. Here, we found that Hes1 oscillation is stable (both the period and amplitude are relatively constant) in the PSM but unstable (the period and amplitude are variable) in the individual dissociated PSM cells. The Hes1 oscillators in the individual dissociated PSM cells are intrinsically similar to those in fibroblasts that are unstable in both period and amplitude. These results indicate that cell-cell communication is essential not only for synchronization but also for stabilization of cellular oscillators. This effect is also simulated by a mathematical model.

Results and Discussion

Because the half life of Hes1 protein is ≈20 min (12), that of the reporter should be 20 min or less. Otherwise, the reporter protein would be accumulated after several cycles of oscillation. In addition, because each peak comes ≈1 h after the trough, the reporter should become active immediately after induction. To overcome these problems, we used the ubiquitinated firefly luciferase as a reporter, which was previously shown to react to such rapid synthesis and degradation processes (20). This luciferase was fused at the N terminus with one (Ub1-Luc) or two (Ub2-Luc) copies of a mutant ubiquitin (G76V) that resist cleavage by ubiquitin hydrolases (20). Both reporter genes were driven by the 2.5-kb Hes1 promoter (Fig. 1a), which contains sites for both Notch induction and negative feedback (Fig. 6, which is published as supporting information on the PNAS web site).

Fig. 1.
Characterization of the Hes1 reporters. (a) Structures of the Hes1 reporters. (b) Bioluminescence of C3H10T1/2 fibroblasts stably transfected with Hes1-Ub1-Luc or Hes1-Ub2-Luc was measured in the presence of cycloheximide (20 μM) (n = 3). Luminescence ...

We first generated clones of C3H10T1/2 fibroblasts stably transfected with each reporter and measured their bioluminescence. Treatment with cycloheximide, a translational inhibitor, showed that both Ub1-Luc and Ub2-Luc proteins were degraded with half lives of ≈10 min and 6 min, respectively, but became stabilized in the presence of the proteasome inhibitor MG132 (Fig. 1b). Luciferase activities of the transfectants carrying either reporter showed at least three cycles of oscillation after serum treatment (Fig. 1c and data not shown), like endogenous Hes1 expression (12). These results indicated that both Hes1-Ub1-Luc and Hes1-Ub2-Luc reporters well mimic the dynamics of endogenous Hes1 expression. Analysis of the time course showed that the peak of luciferase activity came ≈40 min after serum treatment, whereas that of the endogenous Hes1 mRNA came ≈1 h after serum treatment (Fig. 1d). We also compared endogenous Hes1 intron expression, which is detectable only when the gene is actively transcribed, because the intron sequences are present only in nascent transcripts (15, 21). The peak of Hes1 intron expression came ≈40 min after serum treatment (Fig. 1d). Thus, reporter expression was as rapid as endogenous Hes1 intron expression, probably because the reporters do not have any introns and are much shorter in length than the endogenous Hes1 gene. This result also pointed to significant time delays in production of both Hes1 mRNA and Hes1 protein, which were proposed as essential mechanisms for stable oscillatory gene expression (22, 23).

Using the stable transfectants described above, we attempted single-cell imaging of Hes1 expression after serum treatment. Bioluminescence was measured with a highly sensitive cooled charge-coupled device camera, as described (24). A total of 48 cells were monitored, and ≈44% of them (n = 21) were found to exhibit short responses, one or two cycles, and then become almost silenced (Fig. 2 a and b), whereas the others (56%, n = 27) showed longer responses: they were cycling over a 12-h period (Fig. 2 c and d and Movie 1, which is published as supporting information on the PNAS web site). The period and amplitude of each cycle were variable, and the average period was 122 ± 2 min. Strikingly, still many cells showed oscillation 36 h after serum treatment. Most were cycling in an unstable manner, whereas a few were cycling in a relatively stable manner (Fig. 2 e and f and data not shown). Thus, the damped oscillation observed in cultured cells is due not to damped oscillation in all individual cells but to desynchronization between the cycling cells. We also found that, in the absence of serum treatment, many cells were cycling unsynchronously (data not shown), indicating that oscillation goes unnoticed in untreated cultured cells because of desynchronization.

Fig. 2.
Bioluminescence imaging of C3H10T1/2 fibroblasts stably transfected with Hes1-Ub2-Luc. (a) Bioluminescence images of individual cells after serum treatment. Images were taken by 10-min exposure and binning of pixels 8 × 8 to increase signal-to-noise ...

We next generated transgenic mice carrying the Hes1-Ub1-Luc reporter. The caudal part of the embryonic day (E)10.5 transgenic embryo was cultured, and bioluminescence in the PSM was monitored. Hes1 is expressed at a high level in the somite 0 (S0) and at a lower level in the other PSM region (8). Real-time imaging showed that Hes1 oscillation was propagated from the caudal end to S0 in the PSM (Fig. 3a; Fig. 7 and Movie 2, which are published as supporting information on the PNAS web site), and that each cycle generated a pair of somites (Fig. 3c). The caudal region was always earlier in phase than the rostral region (Fig. 3b, compare regions 1 and 2). During the 15-h period, the Hes1 oscillator in the PSM was stable in both period and amplitude (Fig. 3b), with an average period of ≈160 min in our condition. This period was longer than that of Hes1 oscillation and somite segmentation seen in utero (≈120 min). We were reproducibly able to monitor stable Hes1 oscillation and somite segmentation for 15 h, but after that, the growth of the PSM was severely reduced, and the boundaries of the somites became ambiguous, although Hes1 oscillation was still robust and stable even after 1 day.

Fig. 3.
Bioluminescence imaging of the PSM of a Hes1-Ub1-Luc embryo. (a) Bioluminescence images of the PSM were taken by 20-min exposure and binning of pixels 4 × 4 (see Movie 2). Asterisk (top left) indicates S0. Hes1 oscillation was propagated from ...

It was previously shown that expression of the chick homolog c-hairy1 oscillates even in dissected PSM fragments, suggesting that this oscillator functions in a cell-autonomous manner (7, 19). To confirm this result, we dissected the PSM of the Hes1-Ub1-Luc embryo into three fragments and monitored the bioluminescence of each fragment (Fig. 4a). Hes1 oscillation was stable in each fragment for ≈9 h (Fig. 4 b and c), thus confirming the chick results. The caudal-to-rostral phase delay was lost at the first cycle after dissection but soon recovered from the second cycle onward within the same fragment (Fig. 4b; region 1 should be ahead of region 2). In contrast, the relative phase difference between the different fragments was not recovered after dissection (Fig. 4c; region 1 should be ahead of region 3 but became almost the same when separated). Thus, although the dissected PSM fragments maintain a stable Hes1 oscillator, they easily lose their relative phase difference when separated, suggesting that the direct or indirect cell-cell communication is important to keep the precise phase difference.

Fig. 4.
Bioluminescence imaging of the dissected PSM fragments and dissociated PSM cells of Hes1-Ub1-Luc embryos. (a) The caudal part was dissected into three fragments. (b) Quantification of bioluminescence of regions 1 and 2. From the second cycle onward, region ...

It was recently shown that dissociated PSM cells also become out of synchrony (19). However, it is not clear whether each PSM cell has a stable oscillator but is reset at various phases when dissociated or has an unstable oscillator, like fibroblasts. We thus next examined Hes1 oscillation in dissociated PSM cells. Although Hes1 expression oscillated in each dissociated PSM cell (Fig. 4 d and e and Movie 3, which is published as supporting information on the PNAS web site), the period and amplitude were variable. The average period was 155 ± 6 min. Thus, the Hes1 oscillator in most, if not all, individual PSM cells is unstable in period and amplitude; this feature is very similar to that of the Hes1 oscillator in fibroblasts. These results suggest that cell-cell communication may be important not only for synchronization but also for stabilization of cellular oscillators.

To see whether coupling of unstable cellular oscillators would be expected to form stable and synchronized oscillators, we simulated Hes1 oscillation by adapting the model of stochastic coherence with coupling (for details, see Materials and Methods) (25-28). We assumed 128 unstable oscillators, which displayed random oscillatory expression (Fig. 5a and Fig. 8, which is published as supporting information on the PNAS web site). In contrast, coupling by 1D neighboring interaction made all these oscillators stabilized and synchronized (Fig. 5b). These results support the notion that cell-cell communication could not only synchronize but also stabilize the unstable cellular oscillators.

Fig. 5.
Mathematical simulations for uncoupled and coupled oscillators. (a) Mathematical simulation for the uncoupled condition. The time course of 128 unstable oscillators is presented (Left). The trough and peak are shown in black and white, respectively. Three ...

Our results indicate that the Hes1 oscillator of individual PSM cells is unstable and therefore intrinsically similar to that of fibroblasts. This result agrees well with the previous proposal that uncoupling of PSM cells could lead to random fluctuations of the oscillator expression in chick and zebrafish (19, 29). Thus, coupling between cells is very important for an accurate biological clock and may be mediated by Notch signaling in the PSM, where cyclic Hes expression seems to be coupled by cyclic Notch activation (14, 29, 30). Because Notch signaling molecules are expressed in many developing tissues, such tissues might also have a stable and synchronized Hes1 oscillator, which could control accurate timing in development. Our real-time imaging system would offer a powerful tool to look for such oscillating tissues.

Materials and Methods

Plasmids. The reporter constructs Hes1-Ub1-Luc and Hes1-Ub2-Luc were generated as follows. The coding region of one or two copies of ubiquitin G76V was fused to firefly luciferase cDNA in-frame at the 5′ terminus (gift of David Piwnica-Worms, Washington University School of Medicine, St. Louis, MO) (20). The Hes1 promoter and 5′UTR region (-2567 to +223), the Ub1-Luc/Ub2-Luc reporter, the Hes1 3′UTR (+2090 to +2453), and the downstream region (+2454 to +2626) were ligated in this order into pBluescript. The PGK-neo was also cloned at the 5′ end of the Hes1 promoter in reverse orientation for stable transfection.

Cell Culture. C3H10T1/2 fibroblasts were transfected with the linearized Hes1-Ub1-Luc or Hes1-Ub2-Luc plasmid by using the FuGENE6 kit (Roche, Indianapolis), according to the manufacturer's instructions. Neomycin-resistant clones were isolated by standard procedure and maintained in DMEM (GIBCO 11995-065), supplemented with 100 units/ml penicillin/100 μg/ml streptomycin/10% FBS at 37°C in 5% CO2.

Promoter Analysis. Hes1-Ub1-Luc or Hes1-Ub2-Luc (1 μg) was transfected into NIH 3T3 cells, which were plated in six-multiwell plates at a density of 5 × 104 cells/ml, with or without 0.4 μg of the expression vectors for Hes1 or the constitutively active form of Notch. After 48 h, the cells were harvested, and luciferase activities were measured.

Real-Time PCR. Real-time PCR was performed on the ABI PRISM 7300 Sequence Detection System by using SYBR green PCR Master Mix (Applied Biosystems) and 100 nM of each primer. Primer sequences were designed as follows: for Hes1 mRNA, forward primer, 5′-GGACAAACCAAAGACGGCCTCTGAGCACAG-3′; reverse primer, 5′-TGCCGGGAGCTATCTTTCTTAAGTGCATCC-3′; for Hes1 intron expression, forward primer, 5′-AGTTGTTACTGCTCCGGAAATGGAGGGAGA-3′; reverse primer, 5′-CCTGCGGCAGGGGTTGGACCGGTGCTAAAC-3′; for GAPDH, forward primer, 5′-ATCTTCTTGTGCAGTGCCAGCCTCGTCCCG-3′; reverse primer, 5′-AGTTGAGGTCAATGAAGGGGTCGTTGATGG-3′. Cycle conditions were 50°C for 2 min, 95°C for 10 min, and 60 cycles of 95°C for 15 sec and 60°C for 1 min. The specificity of the amplicon was assessed based on the dissociation curve profile. Quantification was determined by the threshold cycle. GAPDH was used as an internal control to normalize the expression.

Transgenic Mice. The 5.1-kb ClaI-NotI DNA fragment, which contains the Hes1 promoter (-2567 to -1), Hes1 5′UTR (+1 to +223), Ub1-Luc, Hes1 3′UTR (+2090 to +2453), and the downstream region (+2454 to +2626), was isolated and injected into the male pronucleus. Genotyping of the Hes1-Ub1-Luc transgenic mice was performed by Southern blot analysis. The region from -374 to +46 of the Hes1 gene was used as a probe. A wild-type band with a size of 6.4 kb and a band of the transgene with a size of 2.8 kb were detected by this probe after digestion with HindIII. All animals used for this study were maintained and handled according to protocols approved by Kyoto University.

Bioluminescence Imaging of C3H10T1/2 Cells. Cells were plated into 35-mm glass-based dishes (ø12-mm glass, IWAKI 3911-035) with 1 ml of DMEM/0.2% FBS/1 mM luciferin (Nacalai Tesque, Kyoto) for 1 day at 37°C in 5% CO2, and the serum was increased to 5-20%. Then the dish was placed on the stage of inverted microscope (Olympus IX81) and maintained at 37°C in 5% CO2. Luminescence from the sample was collected by an Olympus (Tokyo) ×20 UPlanApo objective (numerical aperture 0.8) and transmitted directly to a cooled charge-coupled device camera (Princeton Instruments, Trenton, NJ, VersArray 1 kb). The signal-to-noise ratio was increased by 8 × 8 binning and 10-min exposure.

Bioluminescence Imaging of the PSM. For imaging of the PSM, the caudal part of the embryonic day 10.5 transgenic embryo was transferred into 35-mm glass-based dishes with 200 μl of 100% rat serum with 1 mM luciferin and maintained at 37°C in 5% CO2 and 85% O2. Luminescence from the sample was collected by an Olympus ×20 UPlanApo objective and transmitted directly to a cooled charge-coupled device camera. The signal-to-noise ratio was increased by 4 × 4 binning and 20-min exposure. Under these conditions, we were reproducibly able to monitor the stable Hes1 oscillation and the somite segmentation for 15 h. For imaging of the dissected PSM, we were able to monitor the Hes1 oscillation only for 9 h, because the fragments became degenerated thereafter.

Bioluminescence Imaging of the Dissociated PSM Cells. The PSM without the S0 region was dissected from the embryonic day 9.5-10.5 transgenic embryos and dissociated by brief exposure to trypsin (5 min at 37°C) and mechanical pipetting, as described (19). The dissociated cells were then resuspended in 200 μl of 100% rat serum with 1 mM luciferin to inhibit the trypsin. This cell suspension was plated in 35-mm poly(l-lysine)-coated glass-based dishes and maintained at 37°C in 5% CO2 and 85% O2. Luminescence from the sample was collected by an Olympus ×40 UPLFLN objective (numerical aperture 1.3) and transmitted directly to a cooled charge-coupled device camera. The signal-to-noise ratio was increased by 8 × 8 binning and 10-min exposure.

Image Analysis. Images were analyzed with image-pro plus (Media Cybernetics, Silver Spring, MD). Cosmic ray-induced background noise in the image data was removed. Luminescence intensity was measured within a region of interest defined manually for each cell. The position of the region was adjusted if necessary to accommodate movements of cells during the experiment. Data were logged to Microsoft excel files for plotting and further analysis. Data were corrected for bias by subtracting average luminescence intensity in a no-cell region. When measuring the period of oscillations at the single-cell level, changes with >30% of the biggest peak-trough difference were considered, whereas smaller and abrupt changes (mostly <10% of the biggest) may be due to stochastic fluctuation and were therefore excluded.

Mathematical Modeling. To reproduce the experimental trend by numerical simulation, we adapted the model of stochastic coherence with coupling (25-28). Let ri be a parameter to represent the state of DNA in the ith cell, where ri = 1 and ri = 0 represent the active and inactive states, corresponding to the swelled and compact states, respectively, in the packing of the DNA domain for the related genes (27, 28). Let pi be the number of Hes1 proteins for the ith cell. The rates of the changes of ri and pi are simply expressed as follows.

equation M1

equation M2

equation M3

equation M4

Here, the functions f(ri, pi) and g(ri, pi) represent the local reactions with time constants ε1 and ε2. As the stochastic stimulus, we introduced additive white noises ξi,1 and ξi,2, which are uncorrelated each other (left angle bracketξi,m(tj,n(t′) right angle bracket = δijδmnδ(t - t′), m, n = 1, 2). The third term in the right-hand side of the Eq. 1 represents the neighbor interaction with the strength of k3. Throughout the simulations, we set k1 = 0.20, k2 = 0.40, k4 = 0.050, k5 = 0.75, ε1 = 1/1024, ε2 = 1/2 and D1 = D2 = 2.0 × 10-4. An uncoupled oscillator (k3 = 0) indicates an intermittent behavior (Fig. 8). Fig. 5 shows the time evolutions of the arrays of 128 cells without and with the neighbor interaction. When k3 is set to be zero, each individual cell independently behaves as fluctuating oscillator (Figs. (Figs.5a5a and 8). When k3 is set to be 12, all of the cells oscillate in a synchronized manner (Fig. 5b).

Supplementary Material

Supporting Information:


We thank D. Ish-Horowicz and O. Pourquié for critical reading of the manuscript, D. Piwnica-Worms for the ubiquitin-luciferase reporter, and S. Yamada for production of the transgenic mice. This work was supported by the Genome Network Project and the Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, the Yamanouchi Foundation, and the Uehara Memorial Foundation.


Author contributions: Y.M. and R.K. designed research; Y.M., T.O., Y. Takashima, H.N., Y. Takenaka, and R.K. performed research; Y.M. contributed new reagents/analytic tools; Y.M., T.O., Y. Takashima, H.N., Y. Takenaka, K.Y., H.O., and R.K. analyzed data; and Y.M., H.N., Y. Takenaka, K.Y., and R.K. wrote the paper.

Conflict of interest statement: No conflicts declared.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: PSM, presomitic mesoderm; Sn, somite n.


1. Pourquié, O. (2003) Science 301, 328-330. [PubMed]
2. Bessho, Y. & Kageyama, R. (2003) Curr. Opin. Genet. Dev. 13, 1678-1690.
3. Weinmaster, G. & Kintner, C. (2003) Annu. Rev. Cell Dev. Biol. 19, 367-395. [PubMed]
4. Aulehla, A. & Herrmann, B. G. (2004) Genes Dev. 18, 2060-2067. [PubMed]
5. Giudicelli, F. & Lewis, J. (2004) Curr. Opin. Genet. Dev. 14, 407-414. [PubMed]
6. Rida, P. C. G., Minh, N. L. & Jiang, Y.-J. (2004) Dev. Biol. 265, 2-22. [PubMed]
7. Palmeirim, I., Henrique, D., Ish-Horowicz, D. & Pourquié, O. (1997) Cell 91, 639-648. [PubMed]
8. Jouve, C., Palmeirim, I., Henrique, D., Beckers, J., Gossler, A., Ish-Horowicz, D. & Pourquié, O. (2000) Development (Cambridge, U.K.) 127, 1421-1429. [PubMed]
9. Holley S. A., Geisler, R. & Nüsslein-Volhard, C. (2000) Genes Dev. 14, 1678-1690. [PMC free article] [PubMed]
10. Sawada, A., Fritz, A., Jiang, Y.-J., Yamamoto, A., Yamasu, K., Kuroiwa, A., Saga, Y. & Takeda, H. (2000) Development (Cambridge, U.K.) 127, 1691-1702. [PubMed]
11. Bessho, Y., Sakata, R., Komatsu, S., Shiota, K., Yamada, S. & Kageyama, R. (2001) Genes Dev. 15, 2642-2647. [PMC free article] [PubMed]
12. Hirata, H., Yoshiura, S., Ohtsuka, T., Bessho, Y., Harada, T., Yoshikawa, K. & Kageyama, R. (2002) Science 298, 840-843. [PubMed]
13. Holley, S. A., Julich, D., Rauch, G. J., Geisler, R. & Nüsslein-Volhard, C. (2002) Development (Cambridge, U.K.) 129, 1175-1183. [PubMed]
14. Dale, J. K., Maroto, M., Dequeant, M. L., Malapert, P., McGrew, M. & Pourquié, O. (2003) Nature 421, 275-278. [PubMed]
15. Bessho, Y., Hirata, H., Masamizu, Y. & Kageyama, R. (2003) Genes Dev. 17, 1451-1456. [PMC free article] [PubMed]
16. Hirata, H., Bessho, Y., Kokubu, H., Masamizu, Y., Yamada, S., Lewis, J. & Kageyama, R. (2004) Nat. Genet. 36, 750-754. [PubMed]
17. Kawamura, A., Koshida, S., Hijikata, H., Sakaguchi, T., Kondoh, H. & Takada, S. (2005) Genes Dev. 19, 1156-1161. [PMC free article] [PubMed]
18. Morimoto, M., Takahashi, Y., Endo, M. & Saga, Y. (2005) Nature 435, 354-359. [PubMed]
19. Maroto, M., Dale, J. K., Dequéant, M.-L., Petit, A.-C. & Pourquié, O. (2005) Int. J. Dev. Biol. 49, 309-315. [PubMed]
20. Luker, G. D., Pica, C. M., Song, J., Luker, K. E. & Piwnica-Worms, D. (2003) Nat. Med. 9, 969-973. [PubMed]
21. Morales, A. V., Yasuda, Y. & Ish-Horowicz, D. (2002) Dev. Cell 3, 63-74. [PubMed]
22. Lewis, J. (2003) Curr. Biol. 13, 1398-1408. [PubMed]
23. Monk, N. A. M. (2003) Curr. Biol. 13, 1409-1413. [PubMed]
24. Yamaguchi, S., Isejima, H., Matsuo, T., Okura, R., Yagita, K., Kobayashi, M. & Okamura, H. (2003) Science 302, 1408-1412. [PubMed]
25. Hu, B. & Zhou, C. (2000) Phys. Rev. E 61, R1001-R1004. [PubMed]
26. Han, S. K., Yim, T. G., Postnov, D. E. & Sosnovtseva, O. V. (1999) Phys. Rev. Lett. 83, 1771-1774.
27. Yoshikawa, K. (2002) J. Biol. Phys. 28, 701-712. [PMC free article] [PubMed]
28. Yoshikawa, K. & Yoshikawa, Y. (2002) in Pharmaceutical Perspectives of Nucleic Acid-Based Therapeutics, eds. Mahato, R. I. & Kim, S. W. (Taylor & Francis, London), pp. 138-163.
29. Jiang, Y.-J., Aerne, B. L., Smithers, L., Haddon, C., Ish-Horowicz, D. & Lewis, J. (2000) Nature 408, 475-479. [PubMed]
30. Huppert, S. S., Ilagan, M. X. G., De Strooper, B. & Kopan, R. (2005) Dev. Cell 8, 677-688. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...


  • Compound
    PubChem Compound links
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...