department of basic medical sciences division of molecular ......imai, k., and saito, h. tgf-β...

77

Department of Basic Medical SciencesDivision of Molecular Cell Signaling分子細胞情報分野

1. Regulation of MTK1/MEKK4 kinase activity byits N-terminal autoinhibitory domain andGADD45 binding

Hiroaki Mita1, Junichiro Tsutsui1, MutsuhiroTakekawa, Elizabeth A. Witten1, and Haruo Saito:1Harvard Medical School

In eukaryotes from yeast to mammals, various cel-lular stresses generate intracellular signals thatconverge on the stress-responsive MAP kinases(MAPKs). In mammalian cells, two families of stress-responsive MAPKs, JNK and p38, are activated bystimuli such as UV radiation, ionizing radiation, hy-perosmolarity, oxidative stress, and translationinhibitors, as well as cytokines such as IL-1, TNFα,and TGFβ. These MAPKs are activated by their cog-n a t e M A P K k i n a s e s ( M A P K K s ) t h r o u g hphosphorylation of conserved threonine and ty-rosine residues in the kinase domain activation loop.A MAPKK is activated by specific MAPKK kinases(MAPKKKs) through phosphorylation of conservedthreonine and/or serine residues. An interacting setof MAPKKKs, MAPKKs, and MAPKs constitutes aspecific MAP kinase cascade. Numerous MAPKKKs

are now known to function upstream of the JNK andp38 MAPKs, undoubtedly reflecting the diversestimuli that lead to JNK and p38 activation. While itis believed that different MAPKKKs are activated bydisparate mechanisms, individual mechanisms areeither unknown or only vaguely understood. Hu-man MTK1 (and its mouse homolog MEKK4) is oneof MAPKKKs that act upstream of JNK and p38. Togain more insight into general principles of MAP-KKK activation, we made use of our previousobservation that MTK1 can be expressed functional-ly in yeast cells.

MTK1 has an extensive N-terminal non-catalyticdomain composed of ~1300 amino acids. Full-lengthor near full-length MTK1 is catalytically inactivewhen expressed in yeast cells as is in mammaliancells. Deletion of a segment including position 253-553 activates kinase, indicating that this segmentcontains the auto-inhibitory domain. In the auto-in-hibited conformation, the MTK1 kinase domaincannot interact with its substrate, MKK6. By a func-tional complementation screening using yeast,GADD45 proteins (GADD45α, β, and γ) were identi-fied as MTK1 activators. GADD45 proteins bind asite in MTK1 near the inhibitory domain, and re-

As the global environment deteriorates alarmingly rapidly by pollution, it is becomingcritically vital to learn more about the cellular responses to environmental stressescaused by exposures to, for example, ultraviolet radiation, genotoxins, and oxidants.There is, however, only a rudimentary understanding of the basic mechanisms bywhich cells respond to these environmental stresses. A conspicuous cellular stressresponse is activation of the stress-responsive MAP kinases (JNK and p38), which areconserved throughout the eukaryotic world, indicating their fundamental importancein cellular survival and adaptation. Our primary research goal is to elucidate the molec-ular mechanism of activation of the stress-responsive MAP kinase cascades, usingboth yeast and human cells as model systems.

Professor　 Haruo Saito, Ph.D.Research Associate Mutsuhiro Takekawa, M.D., Ph.D.Research Associate Kazuo Tatebayashi, Ph.D.

教　授理学博士斎　藤　春　雄助　手医学博士武　川　睦　寛助　手薬学博士舘　林　和　夫

78

lieves auto-inhibition. Furthermore, mutants of full-length MTK1 were isolated that can interact withMKK6 in the absence of the activator GADD45 pro-teins. These MTK1 mutants are constitutively active,both in yeast and in mammalian cells.

From these results, we concluded the followings:1) the MTK1 N-terminal region contains an auto-in-hibitory domain that interacts with the C-terminalkinase domain, inhibits kinase activity and preventsinteraction with its substrate MKK6; 2) GADD45 pro-teins are MTK1 activators; and 3) Binding ofGADD45 to the N-terminal region of MTK1 elimi-nates inhibition of the kinase domain by theauto-inhibitory domain. The isolation of constitu-tively active MTK1 mutants, many of which arelocated in the non-catalytic domain, further supportsthese conclusions.

2. Role of the GADD45 proteins in activation ofthe p38 SAPK pathway by TransformingGrowth Factor βββββ

Mutsuhiro Takekawa, Kazuo Tatebayashi, FumioItoh1, Masaaki Adachi1, Kohzoh Imai1 and HaruoSaito: 1First Department of Internal Medicine, Sap-poro Medical University, School of Medicine

Transforming growth factor β (TGF-β) belongs toa family of multifunctional cytokines that regulatecell adhesion, angiogenesis, cell proliferation and ap-optosis. TGF-β expression and responsiveness alsoregulate tumor development. TGF-β initiates itspleiotropic effects by binding a heteromeric cell sur-face receptor complex composed of type I and IItransmembrane S/T kinase receptors. Upon ligandbinding, the type II receptor phosphorylates and ac-tivates the type I receptor. Activated type I receptorinitiates intracellular signaling through activation ofspecific Smad proteins that relay signals into the nu-cleus where they direct transcriptional responses.TGF-β has also been found to activate the stress-re-sponsive p38 MAPK cascade in a variety of cellsystems. The mechanism by which TGF-β activatesthe p38 pathway, however, yet remains to be eluci-dated.

In order to clarify the molecular mechanisms ofTGF-β-induced p38 activation, we first investigatedthe p38 MAPK activity in TGF-β responsive and un-responsive cell lines. We found that TGF-β-inducedp38 activation did not occur in Smad4-deficient celllines, but that reintroduction of Smad4 in these celllines restored the p38 activation. Furthermore, ex-pression of the constitutively active TGFβRIactivated the p38 pathway, and this activation wasfurther enhanced by co-expression of Smad proteins.Perhaps more important, overexpression of Smadproteins alone was capable of activating the p38pathway. In addition, the p38 activation induced bythe constitutively active TGFβRI was strongly inhib-

ited by expression of the dominant-negative Smad4mutant. These findings suggest that Smad-depen-dent gene expression mediates the activation of thep38 pathway in response to TGF-β.

In order to identify the TGF-β-inducible genewhose expression activates the p38 pathway, we in-vestigated expression of the GADD45 familyproteins (GADD45α, β, and γ), which bind to and ac-tivate MTK1 MAPKKK. We found that expression ofGADD45β mRNA was specifically and efficiently in-duced by TGF-β in a Smad-dependent manner, andthat the timing of the TGF-β-induced p38 activationwas almost parallel to that of GADD45β induction.Overexpression of GADD45β was sufficient, even inSmad4-deficient cell lines, to enhance the p38 activi-ty, presumably through the activation of MTK1.Moreover, TGF-β-induced p38 activation wasstrongly inhibited by expression of dominant-nega-tive MTK1 or anti-sense GADD45β. These findingsthus suggest that TGF-β activates the p38 pathway, atleast in part, through Smad-dependent transcriptionof GADD45β.

To clarify physiological roles of TGF-β-inducedp38 activation, we further screened for genes whoseexpression were regulated by both TGF-β-mediatedsignaling and the p38 pathway using a cDNA arraysystem. We identified that expression of thrombo-spondin 1, a potent inhibitor of tumor cell growthand angiogenesis, is induced by TGF-β via Smad-de-pendent p38 activation, suggesting that the p38pathway may play an important role in tumor sup-pression by TGF-β.

3. Usage of tautomycetin, a novel PP1 specificinhibitor, reveals that PP1 is a positive regula-tor of Raf-1 in vivo

Shinya Mitsuhashi1, Hiroshi Shima1, NobuhiroTanuma1, Nobuyasu Matsuura2, MutsuhiroTakekawa, Takeshi Urano3, Tohru Kataoka4, Ma-koto Ubukata2, and Kunimi Kikuchi1: 1Division ofBiochemical Oncology and Immunology, Instituteof Genetic Medicine, Hokkaido University, 2Labo-ratory of Biofunctional Chemistry, BiotechnologyResearch Center, Toyama Prefectural University,3Department of Biochemistry II, Nagoya UniversitySchool of Medicine, 4Division of Molecular Biology,Department of Molecular and Cellular Biology,Kobe University Graduate School of Medicine

Protein phosphatase type1 (PP1), together withprotein phosphatase 2A(PP2A), is a major eukaryoticserine/threonine protein phosphatase (PP) involvedin regulation of numerous cellular functions. Al-though roles of PP2A have been extensively studiedusing okadaic acid (OA), a well-known inhibitor ofPP2A, biological analysis of PP1 has remained re-stricted due to lack of a specific inhibitor. We have

79

recently found that tautomycetin (TC) acts as a high-ly specific inhibitor of PP1 phosphatase. To elucidatethe biological effects of TC, we demonstrated in pre-liminary experiments that treatment of COS-7 cellswith 5mM TC inhibited endogenous PP1 activity bymore than 90% without affecting PP2A activity.Therefore, using TC as a specific PP1-inhibitor, thebiological effect of PP1 on the MAP kinase (MAPK)signaling pathways was examined. First, we foundthat inhibition of PP1 in COS-7 cells by TC selectivelysuppresses activation of ERK, but not stress-respon-

sive MAP kinases (JNK and p38). TC-mediated inhi-bition of PP1 also suppressed activation of Raf-1,resulting in inhibition of MEK-ERK signaling. To fur-ther examine the role of PP1 in regulation of Raf-1,we overexpressed the PP1 catalytic subunit (PP1C) inCOS-7 cells. Ectopic expression of PP1C enhancedRaf-1 kinase activity, whereas a phosphatase-deadPP1C mutant blocked Raf-1 activation in vivo. Fur-thermore, a physical interaction between PP1C andRaf-1 was observed. These data strongly suggest thatPP1 positively regulates Raf-1 in vivo.

Mita, H., Tsutsui, J., Takekawa, M., Witten, E.A., andSaito, H. Regulation of MTK1/MEKK4 kinase ac-tivity by its N-terminal domain and GADD45binding. Mol. Cell. Biol. 22: 4544-4555, 2002

Takekawa, M., Tatebayashi, K., Itoh, F., Adachi, M.,Imai, K., and Saito, H. TGF-β activates p38 MAPKthrough Smad-dependent induction GADD45βexpression. EMBO J. 21: 6473-6482, 2002.

S a i t o H . T h e S l n 1 - Y p d 1 - S s k 1 m u l t i s t e pphosphore lay sys tem that regulates anosmosensing MAP kinase cascade in yeast. In His-tidine Kinases in Signal Transduction. edited byInouye, M. and Dutta, R. (Elsevier Science, San Di-ego), pp397-419, 2003.

Mitsuhashi, S., Shima, H., Tanuma, N., Matsuura, N.,Takekawa, M., Urano, T., Kataoka, T., Ubukata,M.and Kikuchi, K. Usage of tautomycetin, a novelinhibitor of protein phosphatase 1 (PP1), revealsthat PP1 is a positive regulator of Raf-1 in vivo. J.Biol. Chem. 278: 82-88, 2003.

Shonai, T., Adachi, M., Sakata, K., Takekawa, M.,Endo, T., Imai, K. and Hareyama, M. MEK/ERKpathway protects ionizing radiation-induced loss

Publications

of mitochondrial membrane potential and celldeath in lymphocytic leukemia cells. Cell DeathDiffer. 9: 963-971, 2002.

Itoh, M., Adachi, M., Yasui, H., Takekawa, M.,Tanaka, H. and Imai, K. Nuclear export of gluco-corticoid receptor is enhanced by JNK-mediatedphosphorylation. Mol. Endocrinol. 16: 2382-2392,2002.

Matsunaga, T., Ishida, T., Takekawa, M., Nishimura,S., Adachi, M. and Imai, K. Analysis of gene ex-pression during maturation of immature dendriticcells derived from peripheral blood monocytes.Scand. J. Immunol. 56: 593-601, 2002.

Nakamura, K., Tanoue, K., Satoh, T., Takekawa, M.,Watanabe, M., Shima, H., and Kikuchi, K. A novellow-molecular-mass dual-specificity phosphatase,LDP-2, with a naturally occurred substitution thataffects substrate specificity. J. Biochem. 132: 463-470, 2002.

武川睦寛、舘林和夫、斎藤春雄　「ストレス応答MAPキナーゼ」　蛋白質核酸酵素．「MAP キナーゼ研究　現状と展望」　共立出版．47(11): 1379-1387, 2002.

80

Department of Basic Medical SciencesDivision of Neuronal Network神経ネットワーク分野

1. NMDA receptor phosphorylation and synapticplasticity

Shoji Komai1, Ayako M. Watabe, Takanobu Naka-zawa2, Tohru Tezuka2, Tadashi Yamamoto2, andToshiya Manabe: 1Division of Cell Biology andNeurophysiology, Department of Neuroscience,Faculty of Medicine, Kobe University, and 2Divi-sion of Oncology, Department of Cancer Biology

In the hippocampus, excitatory synaptic transmis-sion is regulated dynamically depending on thepattern of synaptic activation: high-frequency activa-tion induces long-lasting enhancement of synapticefficacy referred to as long-term potentiation (LTP),and prolonged lower-frequency activation causeslong-term depression (LTD) of synaptic transmis-sion. Excitatory synaptic transmission is mediated byglutamate receptors and the N-methyl-D-aspartate(NMDA) receptor, one of the glutamate receptor sub-types, plays crucial roles in LTP and LTD induction.

Tyrosine phosphorylation of NMDA receptors bySrc-family tyrosine kinases such as Fyn is implicatedin synaptic plasticity. We identified Fyn-mediatedphosphorylation sites on the GluRε2 (NR2B) subunitof NMDA receptors and Tyr1472 was the majorphosphorylation site. We then generated rabbit poly-

clonal antibodies specific to Tyr1472-phosphorylatedGluRε2, and showed that Tyr1472 of GluRε2 was in-deed phosphorylated in murine brain using theantibodies. Moreover, Tyr1472 phosphorylationgrew evident when mice reached the age when hip-pocampal LTP started to be observed and itsmagnitude became larger. Finally, Tyr1472 phospho-rylation was significantly enhanced after theinduction of LTP in the hippocampal CA1 region.These data suggest that Tyr1472 phosphorylation ofGluRε2 is important for synaptic plasticity. We arecurrently examining mutant mice that have a pointmutation in this residue (tyrosine → phenylalanine)electrophysiologically and behaviorally.

2. Adhesion molecules and synaptic plasticity

Misa Shimuta, Shoji Yamamoto3, Shogo Oka3,Yoichiro Iwakura4, Toshisuke Kawasaki3, andToshiya Manabe: 3Department of Biological Chem-istry, Graduate School of Pharmaceutical Sciences,Kyoto University, and 4Laboratory of Cell Biology,Center for Experimental Medicine, Institute ofMedical Science, University of Tokyo

Adhesion molecules play critical roles in synaptictransmission and plasticity. The HNK-1 carbohy-

Our major research interest is the molecular mechanisms of higher brain functions inmammals such as emotion, and learning and memory. We are especially focusing onthe roles of functional molecules localized in synapses, for instance, neurotransmitterreceptors, signal transduction molecules and adhesion molecules, in neuronal infor-mation processing. We are examining receptor functions, synaptic transmission andplasticity, and their roles in whole animals with electrophysiological, biochemical,molecular genetic and behavioral approaches.

Professor Toshiya Manabe, M.D., Ph.D.Research Associate Ayako M. Watabe, Ph.D.Research Associate Minoru Matsui, M.D., Ph.D.

教　授医学博士　真　鍋　俊　也助　手医学博士　渡　部　文　子助　手医学博士　松　井　　　稔

81

drate epitope, a sulfated glucuronic acid at the non-reducing terminus of glycans, is expressedcharacteristically on a series of cell adhesion mole-cules and is synthesized through a key enzyme,glucuronyltransferase (GlcAT-P). We generatedmice with a targeted deletion of the GlcAT-P gene.The GlcAT-P deficient mice exhibited normal devel-opment of gross anatomical features, but the adultmutant mice exhibited reduced long-term potentia-tion at the Schaffer collateral-CA1 synapses and adefect in spatial memory formation. This is the firstevidence that the loss of a single non-reducing termi-nal carbohydrate residue attenuates higher brainfunctions.

3. Synaptic plasticity at the mossy fiber-CA3synapse

Haruyuki Kamiya1, Kazumasa Umeda1, Seiji Oza-wa5, and Toshiya Manabe: 5Department ofphysiology, Gunma University School of Medicine

a. Kainate receptor-dependent short-term plasticity ofpresynaptic Ca2+ influx

Transmitter release at the hippocampal mossy fi-b e r ( M F ) - C A 3 s y n a p s e e x h i b i t s r o b u s tuse-dependent short-term plasticity with an ex-tremely wide dynamic range. Recent studiesrevealed that presynaptic kainate receptors (KARs),which specifically localized on the MF axons, medi-ate unusually large facilitation at this particularsynapse in concert with the action of residual Ca2+.However, it is currently unclear how activation ofkainate autoreceptors enhances transmitter releasein an activity-dependent manner. Using fluorescencerecordings of presynaptic Ca2+ and voltage in hip-pocampal slices, we demonstrated that paired-pulsestimulation (with 20-200 ms intervals) resulted in fa-cilitation of Ca2+ influx into the MF terminals, asopposed to other synapses, such as the Schaffer col-lateral-CA1 synapse. These observations deviatefrom typical residual Ca2+ hypothesis of facilitation,assuming an equal amount of Ca2+ influx per actionpotential. Pharmacological experiments reveal thatthe facilitation of presynaptic Ca2+ influx is mediatedby activation of KARs. We also found that action po-tentials of MF axons are followed by prominentafterdepolarization, which is partly mediated by ac-tivation of KARs. Notably, the time course of theafterdepolarization approximates to that of thepaired-pulse facilitation of Ca2+ influx, suggestingthat these two processes are closely related to eachother. These results suggest that the novel mecha-nism amplifying presynaptic Ca2+ influx mayunderlie the robust short-term synaptic plasticity atthe MF-CA3 synapse in the hippocampus, and thisprocess is mediated by KARs whose activationevokes prominent afterdepolarization of MF axons

and thereby enhances action potential-driven Ca2+

influx into the presynaptic terminals.

b. Presynaptic Ca2+ entry during mossy fiber LTP

The hippocampal mossy fiber (MF)-CA3 synapseexhibits NMDA receptor-independent long-term po-tentiation (LTP), which is expressed by presynapticmechanisms leading to persistent enhancement oftransmitter release. Recent studies have identifiedseveral molecules that may play an important role inMF-LTP. These include Rab3A, RIM1α, kainate au-toreceptor, and hyperpolarization-activated cationchannel (Ih). However, the precise cellular expressionmechanism remains to be determined because somestudies noticed essential roles of release machinerymolecules, whereas others suggested modulation ofthe ionotropic processes affecting Ca2+ entry into thepresynaptic terminals. Using fluorescence record-ings of presynaptic Ca2+ in hippocampal slices, wedemonstrated that MF-LTP is not accompanied by anincrease in presynaptic Ca2+ influx during an actionpotential. Whole-cell recordings from CA3 neuronsrevealed long-lasting increases in mean frequency,but not mean amplitude, of miniature EPSCs afterthe high-frequency stimulation of MFs. These dataindicate that the presynaptic expression mechanismsresponsible for enhanced transmitter release duringMF-LTP involve persistent modification of presyn-aptic molecular targets residing downstream of Ca2+

entry.

4. Analysis of muscarinic acetylcholine receptorfunctions using knockout mice

Minoru Matsui, Yuji Kiyama, Norikazu Katayama,Fumiko Arima, Toru Shinoe1, Ayako M. Watabe,and Toshiya Manabe

We are investigating the biological function ofmuscarinic acetylcholine receptors (mAChRs) usingmutant mice lacking corresponding genes (mAChRKO mice). These mice have been established by Mat-sui et al. at the Laboratory of Biomedical Genetics,Graduate School of Pharmaceutical Sciences, TheUniversity of Tokyo (Prof. Makoto Mark Taketo’sLab.). The mAChRs (M1, M2, M3, M4 and M5) belongto a group of seven transmembrane-spanning recep-tors and are distributed widely in the both centraland peripheral nervous systems. We have cloned allfive genes for mouse mAChRs (Chrm1, Chrm2,Chrm3, Chrm4, and Chrm5), and determined theirchromosomal locations.

Elucidation of the subtype-specific functions ofmAChRs has been a matter of considerable interests,especially because they are suitable targets for phar-macological therapeutics. However, because of poorsubtype-selectivity of the available ligands, pharma-cological approaches to discriminate their roles

82

remain inconclusive. The use of mAChR KO mice isan alternative strategy to achieve complete subtypespecificity. In order to minimize the concomitant ef-fects reflecting the possible difference in the geneticbackground, we are backcrossing these mice to tworepresentative inbred strains, C57BL/6J and DBA/2J.

We have previously reported that the M3 KO micewere retarded in post-weaning growth and devoid ofpilocarpine-induced salivation. These mice alsoshowed partial mydriasis, and male-selective uri-nary retention. We newly reported the phenotype ofa mutant mouse line that lacks both M2 and M3 re-ceptors (M2/M3 double KO mice). These mice wereviable in spite of complete lack of cholinergic smoothmuscle contraction in vitro. Because these mice didnot develop intestinal obstruction, the widely-ac-cepted view that acetylcholine is essential for normalgut movements such as peristalsis should be chal-lenged. In the eyes, the M2/M3 double KO mice hadsmaller pupils than those of the M3 KO mice, whichsuggests that the signals through M2 and M3 counter-act to control the pupil size. We also found that M2 inthe smooth muscle mediates inhibition of the relax-ant agents that increase cAMP levels.

Finally, mAChRs are supposed to be important invarious brain functions. These include learning andmemory, drug addiction, sleep and respiratory con-trol, and striatal function. We are investigating therole of each subtype in these aspects, employing mo-lecular biology, electrophysiology, and behavioralexperiments. Our mice are now regarded as invalu-able resources and we are organizing manycollaborative programs.

5. Role of Ras/MAP kinase signaling in synapticplasticity

Ayako M. Watabe, Noriko Kumazawa, NorikazuKatayama, Thomas J. O’Dell6, and ToshiyaManabe: 6Department of Physiology, School ofMedicine, University of California, Los Angeles,USA

Small guanine nucleotide-binding protein Ras andmitogen-activated protein kinase (MAPK) signalingcascade has been suggested to play a regulatory rolein the induction of LTP. We therefore examined thechange in synaptic transmission and plasticity in ge-netically manipulated mice that carry no SynGAP, aGTPase-activating protein known to interact withPSD-95 and negatively regulate MAPK signaling.The mutant mice showed a reduced level of LTP atall examined protocols and showed deficits in a hip-pocampus-dependent spatial learning test, whichcan be overcome by excess training. The molecularmechanisms underlying this LTP and learning im-pairment are still unclear, and the studies to identifythe altered MAPK pathways in the mutant mice are

now in progress.

6. Modulatory neurotransmitters and synapticplasticity

Ayako M. Watabe, Hideki Miwa, Fumiko Arima,Thomas J. O’Dell6, and Toshiya Manabe

Several signaling mechanisms that are crucial forthe induction of LTP by theta frequency (5 Hz) trainsof synaptic stimulation are altered in aged animals.Thus, to determine whether the induction of LTP bytheta frequency stimulation is particularly sensitiveto changes in synaptic function that occur in agedanimals, we compared the effects of three differenttrains of synaptic stimulation pulses delivered at 5Hz (theta pulse stimulation, TPS) on synapticstrength in the hippocampal CA1 region of aged andyoung mice. In addition, we investigated whetherthe modulation of TPS-induced LTP by β-adrenergicand cholinergic receptor activation showed deficitswith aging. Our results indicated that TPS-inducedLTP was not diminished in the aged hippocampusbut showed pronounced dependence on L-type cal-cium channels that was not seen in slices from younganimals. In addition, we observed that the enhance-ment of TPS-induced LTP by co-activation ofβ-adrenergic and cholinergic receptors was signifi-cantly reduced in slices obtained from aged animals.Since TPS-induced LTP was not altered in aged mice,our results suggest that deficits in modulatory path-ways that regulate activity-dependent forms ofsynaptic plasticity may contribute to memory im-pairments in older animals. The molecular andbiochemical mechanisms underlying this alterationin aged animals are currently under investigation.

7. Mechanisms of bidirectional synaptic modifi-cation

Ayako M. Watabe, Fumiko Arima, NorikazuKatayama, Hideki Miwa, Noriko Kumazawa, andToshiya Manabe

Activity-dependent modification of synapticstrength plays a key role in neural development andsome forms of neuronal plasticity. While much focushas been on the LTP mechanisms, not much isknown for the molecular mechanisms of LTD, long-lasting suppression of synaptic strength. Recently, ithas been reported that activation of the metabotropicglutamate receptor (mGluR) with the group I mGluRagonist (R,S)-3,5-dihydroxyphenylglycine (DHPG)induces LTD in the CA1 region of the hippocampus.We investigated potential roles of pre- and postsyn-aptic processes in the DHPG-induced LTD.DHPG-induced LTD was completely blocked whenGDP-βS was delivered into postsynaptic cells,strongly suggesting that DHPG depresses synaptic

83

Publications

transmission through a postsynaptic, G protein-me-diated signaling pathway. On the other hand, theeffect of DHPG was strongly modulated by experi-mental manipulations that altered presynapticcalcium influx. Also, enhancing calcium influx byprolonging action potential duration with bath ap-plications of the potassium channel blocker 4-APstrongly reduced the effect of DHPG. Furthermore,while inhibiting both pre- and postsynaptic potassi-um channels with bath-applied 4-AP blocked theeffects of DHPG, inhibition of postsynaptic potassi-um channels alone with intracellular cesium andTEA had no effect on the ability of DHPG to inhibitsynaptic transmission. These results suggest that ac-tivation of postsynaptic mGluRs suppressestransmission at excitatory synapses onto CA1 pyra-midal cells through presynaptic effects ontransmitter release. Further physiological roles of

Yamamoto, S., Oka, S., Inoue, M., Shimuta, M.,Manabe, T., Takahashi, H., Miyamoto, M., Asano,M., Sakagami, J., Sudo, K., Iwakura, Y., Ono, K.and Kawasaki, T. Mice deficient in nervous sys-tem-specific carbohydrate epitope HNK-1 exhibitimpaired synaptic plasticity and spatial learning.J. Biol. Chem. 277:27227-27231, 2002.

Manabe, T. Does BDNF have pre- or postsynaptictargets? Science 295:1651-1652, 2002.

Kamiya, H., Ozawa, S. and Manabe, T. Kainate re-ceptor-dependent short-term plasticity of presyn-aptic Ca2+ influx at the hippocampal mossy fibersynapses. J. Neurosci. 22:9237-9243, 2002.

Kamiya, H., Umeda, K., Ozawa, S. and Manabe, T.Presynaptic Ca2+ entry is unchanged during hip-pocampal mossy fiber long-term potentiation. J.Neurosci. 22:10524-10528, 2002.

Matsui, M., Motomura, D., Fujikawa, T., Jiang, J.,Takahashi, S., Manabe, T. and Taketo, M. M. MiceLacking M2 and M3 Muscarinic acetylcholine re-ceptors are devoid of cholinergic smooth musclecontractions but still viable. J. Neurosci. 22:10627-10632, 2002.

Komiyama, N. H., Watabe, A. M., Carlisle, H. J., Por-ter, K., Charlesworth, P., Monti, J., Strathdee, D. J.C., O’Carroll, C. M., Martin, S. J., Morris, R. G. M.,O’Dell, T. J. and Grant, S. G. N. SynGAP regulatesERK/MAPK signaling, synaptic plasticity and

mGluRs in synaptic transmission and activity-de-pendent modification of synaptic transmission arecurrently under investigation.

While certain patterns of synaptic stimulation canchange synaptic strength, the degree and/or direc-tion of the synaptic modification itself can stronglydepend on the previous history of the synaptic acti-vation. This effect of the history of synaptic activityon synaptic plasticity, or plasticity of synaptic plas-ticity (metaplasticity) has been implicated from thetheoretical point of view in neuronal network devel-opment, but its physiological and biochemicalmechanisms are still unclear. To elucidate molecularand cellular mechanisms underlying metaplasticity,we are examining what kinds of transmitter recep-tors and signaling cascades are involved inmetaplasticity.

learning in the complex with postsynaptic density95 and NMDA receptor. J. Neurosci. 22:9721-9732,2002.

Watabe, A. M., Carlisle, H. J. and O’Dell, T. J.Postsynaptic induction and presynaptic expres-sion of group 1 mGluR-dependent long-term de-pression in the hippocampal CA1 region. J.Neurophysiol. 87:1395-1407, 2002.

Matsui, M., Griffin, M. T., Shehnaz, D., Taketo M. M.and Ehlert F. J. Increased relaxant action offorskolin and isoproterenol against muscarinicagonist-induced contractions in smooth musclefrom M2 receptor knockout mice. J. Pharmacol.Exp. Ther. (in press)

Watabe, A. M. and O’Dell T. J. Age-related changesin theta frequency stimulation-induced long-termpotentiation. Neurobiol. Aging (in press)

松井稔 , 舩田正彦．ムスカリン性受容体サブタイプと疾患－最近の展開－．精神保健研究．48:43-51, 2002.

神谷温之、真鍋俊也．高次脳機能　－学習・記憶と海馬シナプス伝達の可塑性－．総合リハビリテーション．30:915-919, 2002.

真鍋俊也（分担執筆）．海馬におけるシナプス可塑性．脳の発生・分化・可塑性．御子柴克彦、清水孝雄編集．共立出版．166-177, 2002.

神谷温之、真鍋俊也（分担執筆）．神経の可塑性と記憶･学習．分子生物学イラストレイテッド（改訂第2版）．田村隆明、山本雅編集．319-323, 2002.

84

Department of Basic Medical Sciences

Division of Structural Biology (1)分子構造解析分野(生体分子イメージング)

Visualization of structural features of functioningprotein molecules during various molecularevents related to cell motility and signal trans-duction

E. Katayama1 and T. Shiraishi1: 1In collaborationwith Prof. N. Baba’s team, Kogakuin Univ.

We are investigating the three-dimensional (3-D)architecture of various macromolecular assembliesthat might play crucial roles in a number of cell mo-tility and intracellular signal-transduction systems.

In the field of molecular motor research, an origi-nal methodological approach, “single moleculephysiology” was developed 20 years ago and com-pletely innovated the conventional knowledge onthe intrinsic properties of various motor proteins andtheir interactions. A number of amazing results fasci-nated the researchers in other fields and promptedthem to apply new techniques to various fields andmaterials including live cells. The most importantmessage from the new concept is that the behavior ofindividual molecules might be different from eachother and that some new important informationcould be revealed only by measuring unaveragedproperties of each single molecule, separately. Thus,single molecule physiology has already become one

of the most powerful and indispensable approachesin current biophysical sciences. On the other hand,conventional means of structural biology; i.e. X-raycrystallography or multi-dimensional NMR analy-sis, collects the data from a vast number of particlesto be averaged both in time and space, and apparent-ly is not applicable to “single molecule” matter.Electron microscopy is unique in terms that it has apotential to visualize the structure of individualmolecules. In order to obtain the structural informa-tion of functioning actomyosin motor compatible tosingle molecule physiology, we have been utilizingquick-freeze deep-etch replica electron microscopywith mica-flake-technique to capture transient 3-Dconfiguration of myosin crossbridges supporting ac-tin movement in vitro. In this way, a variety ofmolecular events extensively studied under fluores-cence microscope can be instantaneously arrestedwithin one millisecond and the structure of individ-ual protein molecules under well-characterizedexperimental conditions might be clearly visualizedwith a resolution that enables to recognize subdo-main constitution of “individual” protein molecules,by high contrast metal-shadowing. Since replicaspecimens are extremely tolerant to high-dose elec-tron beam irradiation, it should be possible to takemany micrographs of the same field, and to recon-

Electron microscopy provides a useful and unique means to investigate the structureof biological materials including cells/tissues and purified macromolecules. If thespecimens are properly prepared, we can preserve the instantaneous structure offunctioning molecules not only in solution but also in live cells, and visualize theirdetails with high contrast. Though the spatial resolution may not compete with X-raydiffraction, the real superior feature of the method is its almost unlimited applicabilityto those whose structure cannot otherwise be pursued.

Professor Eisaku Katayama, M.D., Ph.D. 教授医学博士片　山　榮　作

85

struct the 3-D structure of individual protein parti-cles by a tomograpic technique, if “missingdata-range problem” can be adequately overcome.We devised a new method along this line (patentpending) and applied it to obtain 3-D image of a sin-gle molecule from tilt-series micrographs of heavymeromyosin (HMM) complexed with ADP/inorgan-ic vanadate (Vi). The fundamental subdomainarrangement of HMM-ADP/Vi in our 3D-image wasvery similar to scallop S1-ADP/Vi structure recentlydetermined by X-ray diffracton, confirming the reli-ability of our new method.

As a complementary approach to characterize the3-D structure of the target particles, we have beensimulating the replica images of protein moleculesfrom their atomic coordinates. Since the image con-trast of replica specimen arises by heavy metalshadowing, virtual model in the shape of proteinparticle was placed in cyber-space, and its image wasrendered by a ray-tracing computer program, as illu-minated by a number of surrounding light-sourcesfrom the appropriate elevation angles (patent pend-ing). We were encouraged by initial trials with themodels of actin filament and actomyosin rigor com-plex, which not only showed generally goodmatching to real replica images, but also revealedsome subtle differences in the configurations in be-tween them, suggestive for the feasibility andusefulness of this unconventional approach. Hence,we applied such strategies to examine 3-D structuralfeatures of myosin heads under a variety of condi-tions; free in solution with various boundnucleotides, and/or associated with actin filament.We had previously found that the body of cross-bridges supporting actin during sliding was oftenbent to hold actin filament inside its curvature. Ac-cording to swinging lever-arm model, however, themotor-domain is presumed to stay attached to actinin the same configuration throughout its workingstroke, and the lever-arm moiety moves along the di-rection almost parallel to actin filament axis. Then,putative actin binding site should come rather out-side the curvature of bent S1 body. Now, theconfiguration we observed very often in actual repli-ca images cannot reconcile with any of the transientstates of swinging lever-arm model. The model alsopostulates that each chemical state of myosin in itsATP-hydrolysis cycle corresponds to certain struc-tural state of the crossbridge configuration. Since theabove crossbridge configuration we observed duringsliding does not seem to match with any one of crys-tal structures so far reported, we sought for thepossibility if myosin head could take such unusualconfiguration only transiently during its ATP-hy-drolysis, and examined the 3-D structure of HMMcomplexed with ADP and various metallo-fluorides.None of them represented the unconventional con-figuration we found. Vertebrate skeletal myosin hastwo highly reactive thiol groups called SH1 and SH2

whose chemical modification greatly affects myo-sin’s intrinsic enzymatic activity. X-ray structure ofS1 revealed that they are actually located at bothends of a single α-helix in the heart of myosin. Fromthis structural feature, it is apparent that they cannotcome close each other as long as that helix is kept sta-ble. It has been reported that these two thiols reacteach other to be cross-linked by certain bifunctionalreagents under limited conditions in the presence ofADP. p-Phenylenedimaleimide (pPDM) is one ofsuch reagents whose span between reactive groups isapproximately 12Å. With electron microscopy, eachhead of pPDM cross-linked HMM/ADP was bentand appeared similar to that of usual Vi-type kinkedhead, at first sight. We noticed, however, that the po-larity of the head curvature was opposite to that wehave observed previously for ATP-bound or ADP/Vi-bound HMM. This cannot be a simple mirror im-age of Vi-type configuration, because the segmentconnecting S1 to S2 appeared strangely crooked ortwisted, whereas smoothly curved Vi-form S1 con-tinued quite naturally to S2 moiety. To characterizepossible new configuration more in detail, we exam-ined the structure of those particles by 3-Dreconstruction and computer-simulation as above.Because our purpose here was to determine whichside of S1 molecule is facing toward outside of thecurvature, we laid our stress on the simulation ofsurface profiles of S1 and compared them with repli-ca images actually observed under electronmicroscope. As the first step, we modified S1 back-bone structure and searched for the candidates inwhich the lever-arm comes closer to the oppositeside of scallop Vi-form. In the structure we chose, thelever-arm orientation was about 90 degrees deviatedfrom the original position. Simulated images of themodified structure placed in appropriatate orien-taion showed nice matching with the delicate surfacefeatures of real replica images of pPDM-HMM. Thus,we might be able to say safely that the actin-bindingsite of pPDM-HMM head faces toward the inner sideof the curvature. Since actin-bound HMM representsessentially similar surface features to that of pPDM-HMM, we concluded that one of the most abundantactin-bound configuration during sliding would be ashort-lived transient state whose SH-loop might bedisrupted. We also have separate spectroscopic evi-dence that pPDM-crosslinked myosin taking aunique configuration. It is apparent that such kind ofinformation might be available only through struc-tural analyses of single molecules. We expect thatour approach “Structural biology of Single Mole-cules” would find fruitful future applications alongthis line.

The other collaborative studies are proceedingmostly on the structural change accompanied withthe function of various motility-related protein sys-tems. These projects include the characterization ofdynein molecules (cytoplasmic and axonemal) in

86

H. Tanaka, K. Homma, A. Hikkoshi-Iwane, E.Katayama, R. Ikebe, J. Saito, T. Yanagida, & M.Ikebe: The motor domain, not a long neck domain,determines the large step of myosin-V: Nature,415: 192-195. 2002

S. Nishikawa, K. Homma, Y. Komori, M. Iwaki, T.Wazawa, A. Hikikoshi Iwane, J. Saito, R. Ikebe, E.Katayama, T. Yanagida, & M. Ikebe: Class VI myo-sin moves processively along actin filamentsbackward with large steps. Biochem Biophys ResCommun. 290: 311-317. 2002

K. Tamano, E. Katayama, T. Toyotome, & C. Sasakawa:Shigella Spa32 is an essential secretory protein forfunctional type lll secretion machinery and unifor-mity of its needle length. J Bacteriol. 184: 1244-1252.2002

S. Yamanaka, E. Katayama, K. Yoshioka, S. Nagaki,M. Yoshida & H. Teraoka: Nucleosome linker pro-teins HMGB1 and histone H1 differentially en-

Publications

search for the origin of force to slide microtubules(with Dr. C. Shingyoji’s team; Dept of Biology, Univ.of Tokyo and Dr. K. Oiwa’s team, Kansai Adv. Res.Ctr,), detailed structures of myosin/metallo-fluoro-nucleotide complex (with Dr. S. Maruta’s team, Soka

Univ.), the srtructure of bacterial flagella and needles(with Dr. C. Sasakawa’s team, Div. of Bacterial Infec-tion in this Institute and Dr. S.-I. Aizawa’s team,CREST.).

hance DNA ligation reactions. Biochem BiophysRes Commun.; 292: 268-273. 2002

M. Fukuda, E. Katayama & K. Mikoshiba: The cal-cium-binding loops of the tandem C2 domains ofsynaptotagmin VII cooperatively mediate cal-cium-dependent oligomerization. J Biol Chem.277: 29315-29320. 2002.

S. Yoshida, E. Katayama, A. Kuwae, H. Mimuro, T.Suzuki & C. Sasakawa: Shigella deliver an effectorprotein to trigger host microtubule destabiliza-tion, which promotes Rac1 activity and efficientbacterial internalization. EMBO J. 21: :2923-2935.2002

M. Tominaga, H. Kojima, E. Yokota, H. Orii, R.Nakamori, E. Katayama, M. Anson, T. Shimmen & K.Oiwa: Higher plant myosin XI moves processively onactin with 35nm steps at high velocity. EMBO J. inpress, 2003.

87

Department of Basic Medical SciencesDivision of Structural Biology (2)分子構造解析分野(微細形態)

1. Genetic and molecular anatomy of the Droso-phila nervous system

Emiko Suzuki

Molecular and genetic studies on the Drosophilanervous system have demonstrated the elaborate ge-netic programs that control the development ofneuronal networks. We are interested in how suchprograms are carried out at the subcellular level.

a. Synapse formation

The neuromuscular junction of Drosophila embry-onic body wall musculature is one of the idealmodels for studying development and function ofsynapses. Each hemisegment of an embryo/larvahas 30 muscle cells, innervated by about 40 motoneu-rons. The neuromuscular projection forms in the lateembryos. During these stages, one can fillet dissectthe embryos and observe or manipulate the develop-ment of neuromuscular networks under a lightmicroscope. By the profound genetic and molecularstudies, several proteins have been identified as syn-aptic target recognition molecules. We are interestedin how these molecules function when pre- and post-synaptic cells encounter. Recently we found that the

postsynaptic muscle cells project filopodia (myopo-dia) from their cell surfaces during the stages whenmotoneurons extend toward their target cells(Ritzenthaler et al., Nature neurosc., 2000). We alsofound that the myopodia interact with the filopodiaof axonal growth cones of their synaptic partners.This interaction is highly specific, and the ectopic ex-pression of target recognition molecules suggestedthat it is controlled by these molecules (Suzuki et al.,J Neurobiol., 2000). We were interested to learn howsuch molecules are expressed on the filopodia andhow they play their roles at the ultrastructural levels.This year, we have developed a new method to ob-s e r v e s u c h f i l o p o d i a l i n t e r a c t i o n s t h r e edimensionally at the resolution of several nm usingimmuno-scanning electron microscopy. Each filopo-dium could be identified by the specific labeling withthe antibodies conjugated with 5 or 10 nm colloidalgold particles. We are now studying the distributionof the synaptic target recognition molecules on thesefilopodia. Our preliminary data suggest that theyplay initial roles for the recognition of synaptic part-ner cells.

b. Phototransduction in the visual system

The phototransduction in Drosophila photorecep-

This laboratory has two major activities. One is to offer various supports for the re-search projects using electron microscopes. This includes the development of newtechniques for electron microscopy. The other activity is our own and collaborativeresearch works on various tissues and cells, mainly of nervous systems, by the combi-nation of fine morphology and molecular biology.

Research Associate Emiko Suzuki, Ph.D.Research Associate Hiroshi Sagara, Ph.D.

助　手医学博士　鈴　木　えみ子助　手医学博士　相　良　　　洋

88

tor cells is a G-protein coupled phosphoinositide(PI)-signaling cascade. Its response is very rapid (re-ceptor potential is generated within ~20 m secondsafter light stimulation) and highly regulated. We areinterested in the regulatory mechanism of this sys-tem. Our recent studies have shown that thetopological coupling of the molecules of the majorsignaling cascades and the related metabolic path-way is essential (Suzuki, in “Atlas of Arthropodsensory receptors” Springer-Verlag,1999). This yearwe focused on the genetic mechanism of the intracel-lular targeting of eye-diacylglycerol kinase (eye-DGK/RdgA protein), that is a key enzyme for theregulation of intracellular level of DG and phospha-tidic acid. In Drosophila phototransduction, DG isthought to regulate the gating of the light-sensitivemembrane channels, and phosphatidic acid is a keyprecursor for the regeneration of membranephophatidyl inositol bisphosphate that is essentialfor the initiation of phototransduction. We havedemonstarted that eye-DGK localizes to the special-ized cellular compartment called subrhabdomericregion, which is close to the photoreceptive organel-lae (rhabdomere) that contain most of the moleculesof phototransduction (rhodopsin, G-protein, phos-pholipase C, light-sensitive channels etc). We wereinterested in how this enzyme is localized to this re-gion. By the expression of partially deletedmolecules of eye-DGK in photoreceptor cells, wefound that the cystein-rich domains (CRDs) in the N-terminal region of eye-DGK is required for theintracelluar localization and the enzyme activity.This is the first demonstration of the function ofCRDs in DGKs in vivo.

2. Cell and developmental biology of the visualsystem

a. Vitamin A metabolism in vertebrates and inverte-brates

Hiroshi Sagara and Emiko Suzuki

Many of the proteins involved in vitamin A me-tabolism are phylogenetically conserved. Amongthese, RPE65 protein, which we previously identifiedin chick retinal pigment epithelial (RPE) cells, showsstriking homology throughout vertebrates, morethan 90% amino acid identities from fish to human,suggesting that this protein has essential roles in vi-tamin A metabolism. Although many researchershave studied this protein, its precise function in vita-min A metabolism is still obscure. For the purpose ofclarifying this issue, we studied the homologue ofRPE65 in Drosophila in which various moleculartools can be applicable. We found that Drosophilahas one orthologue of RPE65, DRPE65. Sequenceanalysis indicated that DRPE65 protein is encoded

by ninaB, one of the phototransduction genes. Re-cently, von Linting and Vogt (2000) showed that thisprotein has β-carotene dioxygenase activity. Thisyear, we analyzed the cellular localization ofDRPE65 by immunohistochemistry. We found that itexists in eyes and digestive tracts. In eyes, it localizedto the cone cells that are thought to contain vitamin Ametabolizing enzymes for visual cycle. In the diges-tive tract, the basal region of the epithelial cells in theoesophagus was immunopositive. These results in-dicate that DRPE65 has functions not only in visualcycle but also in vitamin A metabolism in the diges-tive system.

b. Development of the zebrafish eye

Hiroshi Sagara, Ryo Kurita1 and Ken-ichi Arai1

and Sumiko Watanabe1: 1Division of Molecularand Developmental Biology, Department of BasicMedical Sciences

The zebrafish eye is an excellent model for study-ing development of vertebrate eyes. This year, weisolated a new gene (#61) expressed in the develop-ing eyes. This gene was expressed in the mid-lateralmesoderm in early developmental stages (-12hpf),and became restricted to the choroid fissure region inthe eye and some parts of the brain in later stages(24hpf-). The genetic perturbation of #61 induced thehemorrhage in the eye and brain, and caused the dis-organization of the endothelial cells, suggesting thatthe #61 gene has important roles in the developmentof blood vessels in these regions. Function of thisprotein in endothelial morphogenesis is under inves-tigation.

3. Other collaborative research works

We have done following collaborative researchworks taking mainly the part of morphological anal-yses. As for these projects, please refer to the otherparts of this book.

a. Analysis of the knockout mice of the Clone 22 gene, anovel gene expressed in the nervous system

Hiroshi Sagara, Eishun Muto1, Ken-ichi Arai1 andSumiko Watanabe1

b. Analysis of the pancreatic exocrine cell function inIP3-receptor type 2 and 3 double knockout mice

Hiroshi Sagara, Takeshi Nakamura2 and Katsuhi-ko Mikoshiba3: 2Calcium Oscillation Project,ICORP, JST, 3Division of Molecular Neurobiology,Department of Basic Medical Sciences

89

Suzuki, E. High-resolution scanning electron micros-copy of immunogold-labelled cells by the use ofthin plasma coating of osmium. J. Microscopy.208:153-157, 2002.

Sagara, H., Suzuki, E. and Hirosawa, K. RPE65-re-lated proteins constitute a retinoid and carotenoidmetabolizing protein family throughout the ani-mal and plant kingdom. In The Neural Basis ofEarly Vision. edited by Kaneko, A. (Springer-Verlag, Tokyo). in press.

Noda, T., Sagara, H., Suzuki, E., Takada, A., Kida, H.,and Kawaoka, Y. Ebola virus VP40 drives the for-mation of virus-like filamentous particles with GP.J Virology, 76:4855-4865, 2002.

Kondo, K., Sagara, H., Hirosawa, K., Kaga, K.,

Publications

c. Localization of Kid protein in mitotic cells

Emiko Suzuki, Noriko Nishizumi-Tokai4 andTadashi Yamamoto4: 4Division of Oncology, De-partment of Cancer Biology

d. Molecular and structural analyses of the influenza vi-rus and Ebola virus

Takeshi Noda5, Hiroshi Sagara, Emiko Suzuki andYoshihiro Kawaoka5: 5Division of Virology, De-partment of Microbiology and Immunology

Matsushima, S., Mabuchi, K., Uchimura, H. andWatanabe, T. Hair cell development in vivo and invitro: Analysis by using monoclonal antibody spe-cific to hair cells in the chick inner ear. J. Comp.Neurol. 445: 176-198, 2002.

Kurita, R., Sagara, H., Aoki, Y., Link, B. A., Arai, K.and Watanabe, S. Suppression of lens growth byαA-crystallin promoter-driven expression ofdiphtheria toxin results in disruption of retinal cellorganization in zebrafish. Develop. Biol. in press.

鈴木えみ子、Sarah Ritzenthaler、Deminan Rose、千葉晶．シナプスの標的選択過程におけるシナプス後細胞の挙動：ショウジョウバエでわかったこと．電子顕微鏡、37:212-214, 2002.

90


1. Inositol trisphosphate receptor and Ca2+ sig-naling

Takayuki Michikawa, Mitsuharu Hattori, TakeshiNakamura1, Manabu Yoshida1, Kozo Hamada2,Chihiro Hisatsune2, Akira Futatsugi1, Akinori Ku-ruma2, Hiroko Bannai2, Toshifumi Morimura2,Hirohide Iwasaki2, Tsuyoshi Uchiyama, HiroshiMiyauchi, Wei-Hua Cai, Rei Yokoyama, KeikoUchida, Tomohiro Nakayama, Hideaki Ando,Zhang Songbai, Zhou Hong, Yoko Tateishi, Taka-yasu Higo, Toru Matsu-ura, Jun-ichi Goto, AyukoKurokura, Kazumi Fukatsu, Takafumi Inoue, andKatsuhiko Mikoshiba: 1Calcium OscillationProject, ICORP, JST, 2Laboratory for Develop-mental Neurobiology, Brain Science Institute, TheInstitute of Physical and Chemical Research (RIK-EN)

Division of Molecular Neurobiology(1)脳神経発生・分化分野 (1)

Inositol 1,4,5-trisphosphate (IP3) is a second mes-senger produced through the phosphoinositideturnover in response to many extracellular stimuli(hormones, growth factors, neurotransmitters, neu-trophines, odorants, light, etc.), and controls avariety of Ca2+ -dependent cell functions (cell prolif-eration, differentiation, fertilization, embryonicdevelopment, secretion, muscular contraction, im-mune responses, brain functions, chemical sense,light transduction, etc.) by inducing Ca2+ release(IP3-induced calcium release; IICR) from intracellu-lar Ca2+ store sites such as endoplasmic reticulum(ER) to cytoplasm. IP3 binds to its specific receptor(IP3R) on the Ca2+ store sites. IP3R is an IP3-gatedCa2+ release channel, and could be considered as asignal converter that exchanges the IP3 signal to theCa2+ signal that physiologically acts on a wide vari-ety of targets. Our goal in this research is to elucidatethe structure-function relationship of the IP3R and

We study the molecular mechanisms of expressing diversity and specificity unique tothe nervous system and compare them with those of other tissues, and we undertaketo understand the scenario of the development and differentiation of the nervous sys-tem. We are studying the following subjects by introducing molecular imaging,electrophysiology, molecular and cellular biology.1) We have already identified the importance of IP3 receptor / Ca2+ signaling in fertiliza-tion, cell division and dorsoventral axis formation. We further analyze how Ca2+

oscillation is involved in the brain development and differentiation as well as in thehigher brain function.2) We are studying the role of genes (reelin, cdk5, disabled 1) in neuronal positioningand neurite extension.3) We are studying the role of a gene, (zic), which is involved in neural induction,morphogenesis of the brain, and right and left axis formation.4) In addition, we attempt to express specific genes using virus vectors, and study therole of phospholipids and inositol polyphosphates.

Professor　 Katsuhiko Mikoshiba, M.D., Ph.D.Associate Professor Takafumi Inoue, M.D., Ph.D.Research Associate Takayuki Michikawa, Ph.D.Research Associate Mitsuharu Hattori, Ph.D.

教　授医学博士　御子柴　克　彦助教授医学博士　井　上　貴　文助　手医学博士道　川　貴　章助　手薬学博士　服　部　光　治

91

the physiological roles of IP3R-mediated Ca2+ signal-ing in various cell-types.

We have cloned three types of IP3R (IP3R1, IP3R2,and IP3R3) and have analyzed the structure, func-tion, and expression of each type of IP3R by means ofmolecular biological, biochemical, cell biological,physiological and histochemical approaches. Wehave found that each type has different IP3 binding(e.g., affinity, specificity, Ca2+ sensitivity) and modu-lation (e.g., phosphorylation, calmodulin binding)properties. The IP3R is a polypeptide (~ 2,700 aminoacids) with three major functionally distinct regions:the amino-terminal IP3-binding region, the centralmodulatory region and the carboxy-terminal chan-nel region. Four IP3R subunits assemble to form afunctional IP3-gated Ca2+ release channel, and bothhomo- and heterotetrameric channels are detected.We analyzed the folding structure of the IP3R chan-nel by limited trypsin digestion and have found thatthe IP3R channel is an assembly of four subunits,each of which is constituted by non-covalent interac-t ions of f ive major, well folded structuralcomponents. The IP3 binding core, a minimum es-sential region for specific IP3-binding, resides amongresidues 226-578 of the mouse IP3R1. This year weunveiled three-dimensional structure of the IP3-binding core complexed with IP3 at a resolution of2.2-Å. The asymmetric, boomerang-like structureconsists of an N-terminal β-trefoil domain and a C-terminal α-helical domain containing an ‘armadillorepeat’-like fold. The cleft formed by the two do-mains exposes a cluster of arginine and lysineresidues that coordinate the three phosphorylgroups of IP3. Eleven amino acid residues within theIP3-binding core are involved as IP3-coordinatingresidues.

Ca2+ signaling via IICR often exhibits dynamicchanges in time and space inside a cell (known asCa2+ waves and Ca2+ oscillations). These complexspatiotemporal patterns are not produced by simplediffusion of cytoplasmic Ca2+. The essential ingredi-ents to generate repetitive Ca2+ spikes are positivefeedback, cooperativity, deactivation (including neg-ative feedback) and reactivation. The IP3R1 isregulated by cytoplasmic Ca2+ in a biphasic manner.We found that the positive feedback regulation bycytoplasmic Ca2+ is an intrinsic property of the IP3R1,whereas the negative feedback regulation by Ca2+ ismediated by calmodulin, a ubiquitous and multi-functional Ca2+-dependent regulator protein. Thisfinding suggests that the Ca2+-dependent activationof the IP3R is a fast process, while Ca2+-dependentinactivation is a relatively slow process. This year wefound that Ca2+ induces structural changes in the tet-rameric IP3R purified from mouse cerebellum.Electron microscopy of the IP3R particles revealedtwo distinct structures with 4-fold symmetry: awindmill structure and a square structure. Ca2+ re-versibly promoted a transition from the square to the

windmill with relocations of four peripheral IP3-binding domains. Ca2+ appeared to regulate IP3gating activity through the rearrangement of func-tional domains. We have speculated that thewindmill structure is an active form because highconcentrations of Ca2+ only act as an activator to thepurified IP3R1.

Recently, we identified protein 4.1N as a bindingmolecule for the C-terminal cytoplasmic tail of theIP3R1 using a yeast two-hybrid system. 4.1N andIP3R1 associate in both subconfluent and confluentMadin-Darby canine kidney (MDCK) cells, a wellstudied tight polarized epithelial cell line. In subcon-fluent MDCK cells, 4.1N is distributed in thecytoplasm and the nucleus; IP3R1 is localized in thecytoplasm. In confluent MDCK cells, both 4.1N andIP3R1 are predominantly translocated to the basolat-eral membrane domain; whereas 4.1R, theprototypical homologue of 4.1N, is localized at thetight junctions and other ER marker proteins are stillpresent in the cytoplasm. Moreover, the 4.1N-bind-ing region of IP3R1 is necessary and sufficient for thelocalization of IP3R1 at the basolateral membrane do-main. A fragment of the IP3R1-binding region of4.1N blocks the localization of co-expressed IP3R1 atthe basolateral membrane domain. These data indi-cate that 4.1N is required for IP3R1 translocation tothe basolateral membrane domain in polarizedMDCK cells.

2. Physiological studies of the Ca2+ signaling inCentral Nervous System

Takafumi Inoue, Akinori Kuruma2, Takeshi Naka-mura1, Akira Futatsugi1, Jun-ichi Goto, AyukoKurokura and Katsuhiko Mikoshiba

Analysis of the functional roles of the Ca2+ signal-ing in mammalian brain is one of the most focusedtopics in our research. We have shown that a neu-ronal IP3R1-deficient mouse strain generated bygene-targeting technique exhibits significant reduc-tion of birthrate and abnormal behavior (ataxia andseizure). We found that cerebellar slices preparedfrom IP3R1-deficient mice completely lack long-termdepression (LTD), a model of synaptic plasticity inthe cerebellum. Moreover, a specific antibodyagainst IP3R1, when introduced into wild-typePurkinje cells through patch pipettes, blocked the in-duction of LTD. These data indicate that, in additionto Ca2+ influx through Ca2+ channels on the plasmamembrane, Ca2+ release through IP3R plays an es-sential role in the induction of LTD in Purkinje cells.This year, we have been focusing on detailed Ca2+

dynamics evoked by synaptic activation in Purkinjecell dendrite, and have been revealing Ca2+ concen-tration characteristics in time and space, which willbe a basis for understanding consequence of Ca2+

dynamics and Ca2+ signaling in neural dendrite.

92

In nonneuronal cells such as Xenopus oocytes, IP3is a global messenger that liberates Ca2+ throughoutcytoplasm, but in cerebellar Purkinje cells, repetitiveactivation of parallel fiber-Purkinje cell synapsescauses Ca2+ release that is restricted to individualpostsynaptic local domains. The spatially restrictedIICR might contribute the input specificity of thesynaptic plasticity observed in the parallel fiber-Purkinje cell synapses. Our data suggest that thenegative feedback regulations of Ca2+ mediated bycalmodulin determine the spatial and temporal pat-terns of Ca2+ signaling in cerebellar Purkinje cells bylimiting the amount of Ca2+ being released.

3. Studies on the physiological role of IP3R inbody patterning

Takeo Saneyoshi and Katsuhiko Mikoshiba

The phosphatidylinositol (PI) cycle has been pos-tulated to function in dorso-ventral (D-V) axisformation of many species, as indicated by the actionof lithium. Lithium is assumed to block the recyclingof IP3 into inositol by inhibiting the hydrolysis of in-termediate inositol phosphates. Application oflithium to cleavage stage embryos of Xenopus laevisinduced dorsalization by conversion of ventral me-soderm to dorsal mesoderm, with a concomitantreduction in posterior structures. To determine therole of the PI cycle in patterning the body plan, weisolated functional blocking monoclonal antibodies(mAbs) against the Xenopus IP3R (XIP3R). Ventral in-jection of these monoclonal antibodies at 4-cell stageinduced the formation of a secondary dorsal axis,whereas dorsal injection of the mAbs or normalmouse IgG showed no obvious effects. This impliesthat blockage of IICR in the ventral part of Xenopusembryos converted ventral mesoderm to dorsal me-soderm, thereby generating an ectopic dorsal axis.These results indicate that an active IP3-Ca2+ signal isrequired for ventral differentiation. To ask whichupstream signaling pathways play roles in the ven-tral differentiation, we introduced constitutivelyactive mutant of Gαq , βARK, a Gβq inhibitor, orpanel of inhibitory antibodies against Gαq/11, Gαs/olf, or Gαi/o/t/z in Xenopus embryos. And we ob-tained a conclusion that act ivation of theGαs-coupled receptor relays dorsoventral signal toGβq, which then stimulates PLCβ and then the IP3-Ca2+ system. This year, we extended our study in at-tempts to identify possible candidates downstreammolecules or upstream molecules of IP3-Ca2+ signal-ing as mediating the ventral signal. We havecharacterized several candidates in Xenopus.

4. Studies on the ER dynamics

Hiroko Bannai2, Tomohiro Nakayama, YokoTateishi, Kazumi Fukatsu, Mitsuharu Hattori,

Takafumi Inoue and Katsuhiko Mikoshiba

Recently, ER is regarded as a dynamic organellerather than a static and stable membranous structureas was classically considered. We are characterizingdynamic movement and restructuring of ER in vari-ety of cell types including neurons.

In the neuron, ER is the major membranous com-ponent present throughout the axon. While othermembranous structures such as synaptic vesicles areknown to be transported via fast axonal transport,the dynamics of ER in the axon remained unknown.To elucidate them, we directly visualized the move-ment of two ER-specific membrane proteins, thesarcoplasmic/endoplasmic reticulum calcium-AT-Pase and the IP3R, both of which were tagged withgreen fluorescent protein (GFP) in cultured chickdorsal root ganglion neurons. In contrast to GFP-tagged synaptophysin which moved as vesicles at 1µm/sec predominantly in the anterograde directionin the typical style of fast axonal transport, GFP-tagged ER proteins did not move in a discretevesicular form. Their movement detected by the flu-orescence recovery after photobleaching techniquewas bi-directional, and the rate of the movement wasten-fold slower (~0.1 µm/sec) than fast axonal trans-port and temperature sensitive. The rate ofmovement of ER was also sensitive to low doses ofvinblastine and nocodazole which did not affect therate of synaptophysin-GFP. These results suggestthat ER dynamics in the axon is dependent on the ac-tive transport system in which microtuble and motorproteins are probably involved, but is distinct fromthe well-documented movement of membranousvesicles.

5. Studies on the molecular mechanisms of neu-ral and neural crest formation

J. Aruga2, T.Kitaguchi, Y. Koyabu2, T. Inoue, J.Hoshino, T. Tohmonda, K. Mikoshiba

Zic family found as being expressed abundantly inthe cerebellar granule cell lineage. The Zic genes en-code transcription factor with zinc finger motifs. Wepreviously showed that the genes are vertebrate ho-mologues of Drosophila pair-rule gene, odd-pairedand that the zinc finger motifs of Zic are highly simi-lar to those of Gli family, which has beencharacterized as a factor involved in the Sonic hedge-hog mediated signaling cascade. Furthermore, Zicproteins can bind to the target sequence of Gli pro-teins. Recently, we characterized the Xenopus Zic3 asdetermining the ectodermal cell fate and promote theearliest step of neural and neural crest development.In this year, we revealed that Zic2 and Zic1 co-opera-tively control cerebellar development by regulatingneuronal differentiation. So far, we have shown thatreduced expression of Zic2 in mice results in spina

93

Publications

bifida and holoprosencephaly whereas the disrup-tion of Zic1, a strong homologue of Zic2 that has anoverlapping expression pattern, results in cerebellarmalformation with no apparent abnormalities in theforebrain or in posterior neuropore closure. Mice car-rying one mutated Zic1 allele together with onemutated Zic2 allele (Zic1+/-Zic2+/kd) showed a markedcerebellar folial abnormality similar to, but distinctfrom that found in mice homozygous for the Zic1mutation (Zic1-/-). Abnormalities in the developingZic1+/-Zic2+/kd cerebellum share the following fea-tures with those of the Zic1-/- cerebellum: Apreceding reduction of cell proliferation in the ante-rior external germinal layer, a reduction in cyclin D1expression and enhanced expression of the mitosisinhibitors p27 and p16, and enhancement of Wnt7aexpression. These results indicate that Zic1 and Zic2may have very similar functions in the regulation ofcerebellar development.

6. Studies on the regional specification of thenervous system

J. Aruga2, T.Kitaguchi , Y. Koyabu2, T. Inoue, J.Hoshino, T. Tohmonda, K. Mikoshiba

We also found that the same gene family are in-volved in the regional specification of neural tissuesbesides the role of Zic family in the early stage ofneural development. Gene targeting study of Zicfamilies which are in progress in our laboratoryshowed that some of the Zic genes have essentialroles in the development of dorsal neural tissue in-cluding cerebella. Mutations in human ZIC2 or ZIC3causes congenital anomalies. We investigated thefunctional properties of Zic proteins and their rela-tionship to the GLI proteins. As a molecular basis ofthe Zic-Gli cooperativity, we revealed physical inter-action between Zic and Gli proteins. The zinc fingerdomains of the Zic and Gli proteins physically inter-act and Zic protein enhanced the nucleartranslocalization of Gli proteins. The cooperativitybetween Zic and Gli has been shown by Zic1/Gli3combined mutant, in which Zic1 and Gli3 are shownto act together in the vertebral arch formation . Wealso characterized additional Zic-binding proteins.

7. Studies on the cellular and molecular mecha-nism of neuronal migration and positioning inthe developing brain

T. Ohshima2, K. Saruta2, H. Suzuki2, K. Hayashi,K. Mikoshiba

We have been analyzing several molecules thatare involved in controling neuronal alignment in thebrain. We have shown the critical roles of Reelin, dis-abled-1 (Dab1) and Cdk5/p35 kinase in the positioningof cortical neurons through the analyses of the mu-tant mice. We had demonstrated the synergisticcontribution of Cdk5/p35 and Reelin/Dab1 in the posi-tioning of the cortical neurons. In this year, weanalyzed their roles in the nuclei formation of mousehindbrain and found critical roles of these genes inthe formation of facial nucleus and inferior olive. Wealso studied biochemical relation between Cdk5/p35and Dab1. We found Cdk5/p35 phosphorylates Dab1in vivo.

8. Studies on the role of synaptotagmin

M. Fukuda2, A. Mizutani1, E. Kanno2, C. Saegusa2,Y. Ogata2, T. Kuroda2, F. Hamazato2, K. Mikoshi-ba

We have identified synaptotagmin I or II, a synap-tic vesicle membrane protein, as an inositolpolyphosphate (IP4, IP5, and IP6) binding protein.Synaptotagmin is a family of membrane proteinsthat are thought to be involved in vesicular trafficand it consists of five different domains (an intrave-sicular domain, a transmembrane domain, a spacerdomain, a C2A domain, a C2B domain, and a shortcarboxyl terminus). We have previously shown thatthe distinct roles of the two C2 domains and the shortC terminus of synaptotagmin I in synaptic vesicletraffic. The short C terminus is involved in synapticvesicle docking to presynaptic plasma membranes.The C2A domain regulates vesicular exocytosis bothin neurotransmitter release and neurite outgrowth,whereas the C2B domain is involved in synaptic ves-icle recycling. We have also shown that inositolpolyphosphates block neurotransmitter release bybinding to the C2B domain. However, little is knownabout the function of other domains (e.g. the spacerdomain and the intravesicular domain). In 2001, weexamined the function of the spacer domain andfound that it is involved in membrane localizationsignal. We are now investigating the roles of othersynaptotagmin isoforms in brain function.

Saneyoshi, T., Kume, S., Amasaki, Y. and Mikoshiba,K. The Wnt/Calcium pathway activates NF-ATand promotes ventral cell fate in Xenopus embryos.Nature 417: 295-299, 2002

Bosanac, I., Alattia, J. R., Mal, T. K., Chan, J., Talarico,S., Tong, F. K., Tong, K. I., Yoshikawa, F., Furuichi,T., Iwai, M., Michikawa, T., Mikoshiba, K. andIkura, M. Structure of the inositol 1, 4, 5-

94

trisphoshate receptor binding core in complexwith IP3. Nature 420:696-700, 2002

Nagai, T., Ibata, K., Park, ES., Kubota, M., Mikoshiba,K. and Miyawaki, A. A variant of yellow fluores-cent protein with fast and efficient maturation forcell-biological applications. Nature Biotechnol. 20:87-90, 2002

Suetsugu, S., Hattori, M., Miki, H., Tezuka, T.,Yamamoto, T., Mikoshiba, K. and Takenawa, T.Sustained Activation of N-WASP through Phos-phorylation Is Essential for Neurite Extension De-velopmental Cell 3: 645-658, 2002

Aruga, J., Inoue, T., Hoshino, J. and Mikoshiba, K.Zic2 controls cerebellar development in coopera-tion with zic1. J. Neurosci. 22: 218-225, 2002

Yokoyama, K., Su, IH., Tezuka, T., Yasuda, T.,Mikoshiba, K., Tarakhovsky, A. and Yamamoto, T.BANK regulates BCR-induced calcium mobiliza-tion by promoting tyrosine phosphorylation of IP3receptor. EMBO. J. 21: 83-92, 2002

Zhang, S., Mizutani, A., Hisatsune, C., Higo, T.,Bannai, H., Nakayama, T., Hattori, M. andMikoshiba, K. Protein 4.1N Is Required for Trans-location of Inositol 1,4,5-Trisphosphate ReceptorType 1 to the Basolateral Membrane Domain inPolarized Madin-Darby Canine Kidney Cells.J.Biol.Chem 278: 4048-4056, 2003

Fukuda, M., Kowalchyk, J., Zhang, X., Martin, T. andMikoshiba, K. Synaptotagmin IX regulates Ca2+-dependent secretion in PC12 cells. J. Biol. Chem.277:4601-4604, 2002

Fukuda, M., Kuroda, T. and Mikoshiba, K. Slac2-a/melanophilin, the Missing Link Between Rab27and Myosin Va Implications of a Tripartite ProteinComplex for Melanosome Transport. J. Biol.Chem. 277 (14): 12432-12436, 2002

Fukuda, M., Ogata, Y., Saegusa, C., Kanno, E andMikoshiba, K. Alternative splicing isoforms ofsynaptotagmin VII in the mouse, rat and human.Biochem J. 365: 173-180, 2002

Fukuda,M., Katayama, E. and Mikoshiba, K. The Cal-cium-binding Loops of the Tandem C2 Domainsof Synaptotagmin VII Cooperatively Mediate Cal-cium-dependent Oligomerization. J. Biol. Chem.277(32): 29315-29320, 2002

Hamada, K., Miyata, T., Mayanagi, K., Hirota, J. andMikoshiba, K. Two-state Changes in Inositol 1, 4,5-Trisphosphate Receptor Regulated by Calcium.J. Biol. Chem. 277 (24): 21115-18, 2002

Hayashi,K., Ohshima,T., and Mikoshiba, K. Pak1 IsInvolved in Dendrite Initiation as a DownstreamEffector of Rac1 in Cortical Neurons. Mol.Cell.Neuro 20: 579-594, 2002

Ibata, K., Hashikawa, T., Tsuboi, T., Terakawa, S.,Fengyi Liang, Mizutani, A., Fukuda, M. andMikoshiba,K. Non-polarized distribution ofsynaptotagmin IV in neurons: evidence thatsynaptotagmin IV is not a synaptic vesicle protein.

Neurosci. Res. 43: 401-406, 2002Iwasaki, H., Chiba, K., Uchiyama, T., Yoshikawa, F.,

Suzuki, F., Ikeda, M., Furuichi, T. and Mikoshiba,K.Molecular Characterization of the Starfish Inosi-tol 1,4,5-Trisphosphate Receptor and Its Role dur-ing Oocyte Maturation and Fertilization. J. Biol.Chem. 277: 2763-2772, 2002

Kitaguchi, T., Mizugishi, K., Hatayama, M., Aruga, J.and Mikoshiba,K. Xenopus Brachyury regulatesmesodermal expression of Zic3, a gene controllingleft-right asymmetry. Develop. Growth Differ 44:55-61, 2002

Konishi, Y., Matsu-ura, T., Mikoshiba, K. andTamura, T. Basic Helix-Loop-Helix (bHLH) Tran-scription Factors in the Nervous System. Curr.Topics in Neurochem. (in press) 2002

M.Salanova,G. Priori, V. Barone, E. Intravaia, B.Flucher, F. Ciruela, R.A.J. Mcllhinney, J.B. Parys,Mikoshiba,K. and V. Sorrentino Homer proteinsand IP3 receptors co-localise in the longitudinalsarcoplasmic reticulum of skeletal muscle fibres.Cell Calcium 32(4): 193-200, 2002

Matsu-ura, T., Nakadai, T., Oda, Y., Nagasu, T.,Mikoshiba, K. and Tamura, T. Seizure-mediatedaccumulat ion of theβsubuni t or Ca 2+ /calmodulin-dependent protein kinase II in nucleiof mouse brain cells. Neurosci. Lett. 322 :149-152,2002

Ohshima, T. and Mikoshiba, K. Reelin Signaling andCdk5 in the Control of Neuronal Positioning Mo-lecular Neurobiology 26(2-3): 153-166, 2002

Ohshima, T., Ogawa, M., Takeuchi, K., Takahashi, S.,Kulkarni, A. and Mikoshiba, K. Cdk5/p35 contrib-utes synergistically with Reelin/Dab1 to the posi-tioning of facial branchiomotor and inferior oliveneurons in the developing mouse hindbrain. J.Neurosci. 22(10): 4036-4044, 2002

Saegusa, C., Fukuda, M., and Mikoshiba, K.Synaptotagmin V Is Targeted to Dense-coreVesicles That Undergo Caalcium-dependent Exo-cytosis in PC12 Cells. J. Biol. Chem. 277 (27): 24499-24505, 2002

Sheng-Tian, Li., Kato, K., Tomizawa, K., Matsushita,M., Moriwaki, A., Matsui, H. and Mikoshiba, K.Calcineurin Plays Different Roles in Group IIMetabotropic Glutamate Receptor- and NMDAReceptor-Dependent Long-Term Depression. J.Neurosci. 22(12):5034-5041,2002

Tamura, T. , Matsu-ura, T. , Konishi, Y. andMikoshiba, K. Neural Activity-Inducible Tran-scription Factors: Identification of Ca2+ -RelatedTranscriptional Regulators Induced in the MouseBrainFollowing Pentylenetetrazol-Mediated Sei-zure. Recent Res Devel. Mol. Cell. Biol. 3: 75-86,2002

Taruho,S.Kuroda, Fukuda,M. Ariga,H. ,andMikoshiba,K. The Slp Homology Domain ofSynaptotagmin-like Proteins 1-4 and Slac2 Func-

95

tions as a Novel Rab27A Binding Domain. J. Biol.Chem. 9212-9218, 2002

Taruho, S. Kuroda, Fukuda, M., Ariga, H. andMikoshiba,K. Synaptotagmin-like protein 5: anovel Rab27A effector with C-terminal tandem C2domains. Biochem. Biophys. Res. Commun. 293:899-906, 2002

Uchiyama, T., Yoshikawa, F., Hishida, A., Furuichi,T. and Mikoshiba, K. Anovel recombinant hyper-affinity Insitol 1,4,5-trisphosphate (IP3) absorbenttraps IP3, resulting in specific inhibition of IP3 –

mediated calcium signaling. J. Biol. Chem. 277:8106-8113, 2002

Saneyoshi,T., Kume, S. and Mikoshiba,K. Calcium/Calmodulin-Dependent Protein Kinase I in Xeno-pus laevis. Comparative Biochemistry and physiol-ogy (in press)

Ando, H., Mizutani, A., Matsu-ura, T. and Mikoshiba,K. IRBIT, a Novel Inositol 1, 4, 5-Trisphosphate(IP3)Receptor Binding Protein, Is Released from the IP3Receptor upon IP3 Binding to the Receptor. J. Biol.Chem. (in press)

96


Division of Molecular Neurobiology(2)脳神経発生・分化分野 (2)

1. Full-length-enriched cDNA library

As of October 2002, the entire genomic nucleicacid sequence of Plasmodium falciparum (23 Mb on 14chromosomes) had been deciphired, and 5286 geneshad been predicted. However, analyses of mRNAare mandatory for accurate characterization of theexpressed genes. In a cooperative study with Dr. Su-mio Sugano, a full-length-enriched cDNA librarywas produced from the erythrocyte-stage malariaparasites. 5' end-one-pass-sequencing of randomclones provided important information which com-plements the genome sequencing projects. We aredeveloping a mapping viewer that visualizes the se-quences of cDNA clones on the determined genome.The database “FULL-malaria” is now available athttp://fullmal.ims.u-tokyo.ac.jp.

2. Development of a novel DNA vaccine againstthe malaria parasite

Using a full-length cDNA library that was con-structed using the RNA from erythrocyte-stage

Malaria kills more than two million people worldwide every year, most of whom arechildren under 5 years of age. This death toll makes it imperative to develop effectivemethods to control this disease. To achieve this goal, we are focusing on genomeresearch, DNA vaccine development and investigation of chaperones of malaria para-sites.

parasites of murine malaria and an expression vec-tor, we have started the screening of potential DNAvaccines in a murine malaria model.

3. Chaperone DnaJ homologues of malaria para-sites

DnaJ was first described by researchers at our in-stitute as a gene that regulates phage replication in E.coli. In the genomic sequence of Plasmodium falci-parum, 53 species of DnaJ homologues, which arecharacterized by a so-called J domain consisting of 70conserved amino acids, have been identified.Though their ubiquitous existence in all organismsindicates the importance of these molecules, little isknown about their functions. Malaria parasites con-tain quite unique DnaJ homologues (RESA;ring-infected erythrocyte surface antigens). We havefocused on Pfj2 (Pbj2 in murine malaria), which is lo-calized in the ER and unique to Apicomplexaspecies. Systematic analyses will reveal the exactfunctions of these DnaJ homologues in parasitism.

Research Associate Junichi Watanabe, M.D., Ph.D. 助　手医学博士　渡　邉　純　一

97

Junichi Watanabe, Masahide Sasaki, Yutaka Suzuki,Sumio Sugano. Analysis of transcriptomes of hu-man malaria parasite Plasmodium falciparum usingfull-length enriched library: identification of novelgenes and diverse transcription start sites of mes-

Publications

senger RNAs. Gene 291, 105-113, 2002.Junichi Watanabe, Masahide Sasaki, Yutaka Suzuki,

Sumio Sugano. Full-Malaria in The Molecular Bi-ology Database Collection: 2002 update AndreasD. Baxevanis. Nucleic Acids Res. 30, 1-12, 2002.

98


Division of Molecular Biology(1)遺伝子動態分野 (1)

1. Translational Control and Protein-tRNA Mo-lecular Mimicry

Koichi Ito, Makiko Uno1, Kuniyasu Yoshimura2,Miki Wada, Hanae Sato, Norbert Polacek3, MariaJ. Gomez3, Liqun Xiong3, Alexander Mankin3, andYoshikazu Nakamura: 1Biological Research Labo-ratories, Sankyo Co. LTD., 2Department ofBiochemistry II, Jikei University, School of Medi-c i n e , a n d 3C e n t e r f o r P h a r m a c e u t i c a lBiotechnology, University of Illinois

Termination of protein synthesis takes place onthe ribosomes as a response to a stop, rather than asense, codon in the ‘decoding’ site (A site). Transla-tion termination requires two classes of polypeptiderelease factors (RFs): a class-I factor, codon-specificRFs (RF1 and RF2 in prokaryotes; eRF1 in eukary-otes), and a class-II factor, non-specific RFs (RF3 inprokaryotes; eRF3 in eukaryotes) that bind guaninenucleotides and stimulate class-I RF activity. Theunderlying mechanism for translation terminationrepresents a long-standing coding problem of con-siderable interest since it entails protein-RNA

recognition instead of the well-understood codon-anticodon pairing during the mRNA-tRNAinteraction.

a. Making sense of mimic in translation termination

Recent crystallographic evidence suggests that theeukaryotic release factor (eRF1), the bacterial releasefactor (RF2) and the ribosome recycling factor (RRF)all mimic a tRNA shape, while biochemical and ge-netic evidence supports the idea of a tripeptide‘anticodon’ in bacterial release factors RF1 and RF2.However, the suggested structural mimicry of RF2 isnot in agreement with the tripeptide ‘anticodon’ hy-pothesis and furthermore very recent structuresdetermined by cryo-electron microscopy show thatRF2 has a conformation on the ribosome that is dis-tinct from the RF2 crystal structure. Also, hydroxylradical probings of RRF on the ribosome is not inagreement with the simply idea of RRF mimicking atRNA in the ribosome A site. All of this evidence se-riously question a simple concept of mimicry inshape between proteins and RNA, and thus leavesonly mimicry in function of protein factors of transla-tion to be studied.

Mimicry is a sophisticate program developed in animal, fish or plant to cheat objects byimitating a shape or a color for diverse purposes such as to prey, evade, lure, pollinateor threaten, but this is not restricted to a ‘macro-world’ but can be extended to a ‘mi-cro-world’ as ‘molecular mimicry’. Recent advances in the structural and molecularbiology uncovered that a set of translation factors resembles a tRNA shape and, in onecase, even mimics a tRNA function for deciphering the genetic code. Nature must haveevolved this ‘art’ of molecular mimicry between protein and ribonucleic acid usingdifferent protein architectures to fulfill the requirement of the ribosome. The mecha-nism of translational control as well as the structural, functional and applied aspects ofmolecular mimicry are main research interests in this department.

Professor Yoshikazu Nakamura, Ph.D.Research Associate Koichi Ito, Ph.D.Research Associate Akihiko Oguro, Ph.D.

教　授理学博士　中　村　義　一助　手理学博士伊　藤　耕　一助　手理学博士小　黒　明　広

99

b. The critical role of the universally conserved A2602 of23S ribosomal RNA in the release of the nascent pep-tide during translation termination

The ribosomal peptidyl transferase center is re-sponsible for two fundamental reactions, peptidebond formation and nascent peptide release, duringthe elongation and termination phases of proteinsynthesis, respectively. We used in vitro genetics toinvestigate the functional importance of conserved23S rRNA nucleotides located in the peptidyl trans-ferase active site for transpeptidation andpeptidyl-tRNA hydrolysis. While mutations atA2451, U2585, and C2063 (E. coli numbering) did notsignificantly affect either of the reactions, substitu-tion of A2602 with C or its deletion abolished theribosome ability to promote peptide release but hadlittle effect on transpeptidation. This indicates thatthe mechanism of peptide release is distinct fromthat of peptide bond formation, with A2602 playinga critical role in peptide release during translationtermination.

c. Polypeptide release at sense and non-cognate stopcodons by localized charge-exchange alterations intranslational release factors

Only recently has it been established that a tripep-tide in the bacterial release factors, RF1 and RF2,serves as the ‘anticodon’ in deciphering stop codonsin mRNA. However, the molecular basis of the accu-racy of stop codon recognition is unknown.Although specific tripeptides in the RFs are primari-ly responsible for selective reading of cognate stopcodons, charge-flip variant RF proteins, altered atconserved Glu residues adjacent to the tripeptide‘anticodon’, are shown here to be crucial to codonrecognition. Changes of these Glu residues are capa-ble of triggering polypeptide release at non-cognatestop codons, and also at sense codons. These changesalso reverse the growth inhibition by RFs containing‘harmful’ tripeptide-anticodon changes. These find-ings suggest that electrostatic interactions involvingnegative charges in domain C of the release factorsmediate their accurate docking in the ribosome. Ourresults also establish that the charge flipping createsa novel phenotype - translation termination by‘codon bypassing’ via relaxed positioning of the RFtripeptide anticodon in the decoding pocket of the ri-bosome.

d. Omnipotent decoding potential resides in eukaryotictranslation termination factor eRF1 of variant-codeorganisms and is modulated by the interactions ofamino acid sequences within the domain 1

In eukaryotes, a single translational release factor,eRF1, deciphers three stop codons, though its decod-ing mechanism remains puzzling. In the ciliate

Tetrahymena thermophila, UAA and UAG codons arereassigned to glutamine codons. A yeast eRF1-do-main swap containing Tetrahymena domain 1responded only to UGA in vitro and failed to comple-ment a defect in yeast eRF1 in vivo at 37 ºC. Thisdemonstrates that decoding specificity of eRF1 fromvariant code organisms resides at the domain 1.However, the wild-type eRF1 hybrid fully restoredthe growth of eRF1-deficient yeast at 30 ºC. Tetrahy-mena eRF1 contains a variant sequence, KATNIKD,at the tip of domain 1. The TASNIKD variant of hy-brid eRF1 rendered the eRF1-nullified yeast viable,while in an in vitro assay the same hybrid eRF1 re-sponses only to UGA. Nevertheless, the yeast eRF1bearing KATNIKD motif instead of TASNIKS hep-tapeptide present in higher eukaryotes remainsomnipotent in vivo. Collectively, these data suggestthat variant genetic code organisms like Tetrahymenahave an intrinsic potential to decode three stopcodons in vivo and that interaction within the domain1 between the KAT tripeptide and other sequencesmodulates the decoding specificity of TetrahymenaeRF1.

e. Cloning and sequence analysis of translational re-lease factors eRF1 and eRF3 of Pneumocystis carinii

To clarify the translation termination apparatus ofPneumocystis carinii, we searched for structural genesfor eRF1 and eRF3 in P. carinii EST database, andfound an eRF1 homolog as well as an eRF3 homologwhose C-terminal part is highly homologous to EF-1alpha. Based on these sequences, full-length eRF1and eRF3 cDNAs were cloned by the RACE method.eRF1 is composed of 432 amino acids and 80 % and78% similar to those of S. pombe and S. cerevisiae, re-spectively. eRF3 is 629 amino acid long, and containsunique N-terminal sequence, while the C-terminalsequence is homologous to other eukaryotic eRF3.The activity of these homologs was further investi-gated by the in vivo complementation test usingtemperature-sensitive eRF1 (sup45) and eRF3 (sup35)genes of Saccharomyces cerevisiae under heterologousconditions. P. carinii eRF1 gene restored the growthof ts sup45 cells while the eRF3 gene failed to restorethe sup35 defect.

The genomic sequences for eRF1 and eRF3 wereamplified from P. carinii DNA by the PCR methodusing primer sequences of the cDNA end sequences.The deduced eRF1 and eRF3 genes are composed of1884 and 2323 nucleotides, respectively. Surprising-ly, when compared with their cDNAs, both genesappeared to contain unusually multiple introns, i.e.,13 (Pc-eRF1) and 10 (Pc-eRF3) introns. To our knowl-edge, this feature is unique among phylogeneticallyrelated release factors. Most of yeast and funguseRF3s do not encode an intron and S. pombe eRF3may be the rare exception as it encodes one intron.All introns of Pc-eRF1 and Pc-eRF3 are similar in the

100

sizes (39-52 bp) and retain the acceptor/donor siteconsensus sequence for splicing, GT/AG, but do notconserve the apparent branch point consensus asfound in other genes. Functional study of P. cariniirelease factors is in progress in this laboratory.

2. Regulation of Ribosome Recycling

Toshinobu Fujiwara, Koichi Ito, Tohru Yamami,Yuya Watanabe, Tomohiko Toyoda, Brian K.Blakea4, Steven L. Alama4 and Yoshikazu Nakamu-ra: 4Department of Biochemistry, University ofUtah

After release of nascent polypeptides, the postter-mination complex composed of the ribosome,deacylated tRNA, RF and mRNA needs to be dissoci-ated for the next round of protein synthesis.Ribosome recycling factor (RRF), in concert withelongation factor EF-G, is required for disassemblyof the posttermination complex. Three groups in-cluding my group have recently solved the crystalstructure of RRF. These three molecules are com-posed of two domains, domain 1 and domain 2,bridged by two loops (a hinge), and superimposewith tRNA except for the amino acid-binding 3' end.It has been proposed that RRF is a near perfect tRNAmimic to explain the mechanistic disassembly of theposttermination ribosomal complex. RRF, however,is architecturally different from tRNA in that thehinge of RRF forms a flexible ‘gooseneck’ elbow,while the elbow of tRNA is rigid, and that this flexi-bility of RRF is vital for its function. Moreover, thebiochemical findings show that RRF and EF-G splitthe ribosome into subunits in a reaction that requiresGTP hydrolysis. Therefore, the mechanistic signifi-cance of tRNA mimicry by RRF remains to beverified. We assume that nature may not have creat-ed such protein of a tRNA mimic to simply substitutefor tRNA unless protein is required to pursue somefunction(s) that tRNA cannot do.

a. Elongation factor G participates in ribosome disas-sembly by interacting with ribosome recycling factorat their tRNA-mimicry domains

Elongation factor G is a G-protein with motorfunction that drives two target molecules, a tRNA inthe translating ribosome and the ribosome recyclingfactor (RRF) in the post-termination complex. How Gprotein motor action is transmitted to RRF is un-known. Thermus thermophilus RRF is nonfunctional inEscherichia coli. It became functional upon introduc-ing a plasmid expressing E. coli EF-G with surfacechanges in its tRNA-mimic domain, or by replacingthe E. coli EF-G tRNA-mimic domain by the Thermusdomain. Thermus RRF could also be activated by in-troducing surface substitutions in its anticodonarm-mimic region. These ‘gain-of-function’ pheno-

types depend on the combination of heterologousEF-G and RRF alleles. These mutational studies sug-gest that EF-G motor action is transmitted to RRF byspecific surface contacts between the domains thatmimic the anticodon arm.

b. In vitro dissection of ribosomal occupancy of ribo-some recycling factor and elongation factor G

We have previously reported a simple in vitrospin-down system to assay the ribosome-binding ca-pacity of variant RRF proteins. We have extendedthis system to dissect, in some detail, the ribosomaloccupancy of RRF in relation to other translation fac-tors and antibiotics. The data point out the shared aswell as distinct binding sites on the ribosome for RRFand other components such as EF-G.

c. Backbone 1H, 13C, and 15N assignments of the ribo-some recycling factor from Thermus thermophillus

Although structures of RRFs are known, little isknown about the structural features that are respon-sible for the recycling action. We have previouslysuggested by genetic studies that ribosome recyclingis dependent upon the structural arrangement or dy-namic properties of the linker, or hinge regions,between the two domains. Here we report backbone1H, 15N and 13C resonance assignments for the RRFfrom Thermus thermophillus that were obtained frommulti-dimensional triple-resonance NMR tech-niques. The resonance assignments are paramountfor structural and dynamic comparisons of RRFs (ofdiffering organisms) and genetic variants within thehinge region that modulate RRF function.

3. Molecular Biology of Yeast Prions

Toru Nakayashiki, Colin G. Crist, Hideyuki Hara,Hiroshi Kurahashi, Hiroyuki Kodama andYoshikazu Nakamura

The Sup35 protein of the budding yeast Saccharo-myces cerevisiae is a subunit of the eukaryoticpolypeptide-release factor (eRF3) and is essential forterminating protein synthesis at stop codons. Sup35palso exists as a stable amyloid fibril, termed [PSI+],that propagates its aberrant fold in the cytoplasm in amanner analogous to the “protein only” transmis-sion of mammalian prion protein (PrP). [PSI+] cellsare marked by an altered protein conformation ofSup35p whereby the protein is converted from a sol-uble, active state to an aggregated inactive state. Inthe aggregated state, ribosomes often fail to releasepolypeptides at stop codons, causing a non-Mende-lian trait easily detected by the suppression ofnonsense mutations. Thus, the conversion of solubleSup35p [psi-] to the aggregated form [PSI+] serves asa useful model for studying the formation of amyloid

101

deposits and the prion-like transmission of an al-tered protein conformation.

a. [PHI+], a novel Sup35-prion variant propagated withnon-Gln/Asn oligopeptide repeats in the absence ofthe chaperone protein Hsp104

The [PSI+] element of Saccharomyces cerevisiae is anaggregated form of the translation release factorSup35p that is propagated and transmitted cytoplas-mically in a manner analogous to that of mammalianprions. The N-terminal of Sup35p, necessary for[PSI+], contains oligopeptide repeats and multipleGln/Asn residues. We replaced the Gln/Asn-richprion repeats of Sup35p with non-Gln/Asn repeatsfrom heterologous yeast strains. These non-Gln/Asnrepeat Sup35ps propagated a novel [PSI+] variant,[PHI+], that appeared de novo 103 times more fre-quent than [PSI+]. [PHI+] was stably inherited in anon-Mendelian fashion, but not eliminated upon theinactivation of Hsp104p, unlike known [PSI+] ele-ments. In vitro, non-Gln/Asn repeat domains formedamyloid fibers that were shorter and grew moreslowly than did Gln/Asn-rich prion domains, while[PHI+] aggregates were smaller than [PSI+] aggre-gates in vivo. These findings suggest the existence ofan alternative, Hsp104-independent pathway to rep-licate non-Gln/Asn variant Sup35 prion seeds.

b. Species signature elements of Sup35p [PSI+] prions:N-terminal discriminator and peptide-repeat function

The N-terminal of Sup35p, necessary for [PSI+],contains two species-signature elements: N-terminalGln/Asn-rich sequence and the following oligopep-tide-repeats. We showed that Saccharomyces cerevisiae[PSI+] is efficiently transmitted to heterotypicSup35ps that substituted residues 1-41 of S. cerevisiaeSup35p in vivo and in vitro, and that cross-seeded[PSI+] elements not only form [PSI+]-like coaggre-gates with S. cerevisiae [PSI+] seed but are stablyinherited in the absence of the seed. This coaggrega-tion-based apparent, not genuine, [PSI+] state canalso be induced with heterotypic Sup35ps by replac-ing the peptide-repeat signature with the other.These findings suggest that N-terminal Gln/Asn-r ich and peptide-repeat e lements interactindependently with their homologous sequences forspecies discrimination and fibril propagation, re-spectively, and that two signature elements arenecessary for propagating a heritable [PSI+] elementin yeast.

4. Protein Crystallography

Tomohiko Toyoda, Koichi Ito, Masayo Urata, Mar-ia B. Garber5, Natalia Nevskayaa5, StanislavNikonova5, and Yoshikazu Nakamura: 5Protein Re-

search Institute, Pushchino, Russia

a. Structure of ribosomal protein L1 from Methanococ-cus thermolithotrophicus. Functionally importantstructural invariants on L1 surface

The crystal structure of ribosomal protein L1 fromthe archaeon Methanococcus thermolithotrophicus hasbeen determined at 2.8Å resolution. The crystals be-long to space group P212121, with unit-cell parametersa = 67.4Å, b = 70.0Å, c =106.2Å and two molecules perasymmetric unit. The structure was solved by the mo-lecular-replacement method with AmoRe and refinedwith X-PLOR to an R value of 20.7% and an R free of30.5% in the resolution range 8-2.8 Comparison of thisstructure with those obtained earlier for two L1 pro-teins from other sources (the bacterium Thermusthermophilus and the archaeon Methanococcus jann-aschii) as well as detailed analysis of intermolecularcontacts in corresponding L1 crystals reveal structuralinvariants on the molecular surface, which are proba-bly important for binding the 23S ribosomal RNA andthe protein function within the ribosome.

5. Pharmaceutical RNA Discovery by SELEX

Takashi Ohtsu, Akihiro Oguro, Taichi Sakamoto,Kiyotaka Mochizuki, Eiko Futami-Takada, YukikoIwata, Katsushi Koda and Yoshikazu Nakamura

The systematic evolution of ligands by exponen-tial enrichment (SELEX) method is based on the invitro selection of oligo-nucleotide ligands from largerandom-sequence libraries by repeated reactions ofDNA transcription, RNA selection and RT-PCR am-plification. The selected oligo-nucleotide ligands arecalled ‘aptamer’ which has high affinity and specific-ity to target molecules. We have initiated SELEXexperiments using mammalian translation initiationfactors including eIF4G and eIF4A provided by Dr.Nahum Sonenberg (McGill University, Canada).eIF4G and eIF4A proteins are known to be crucial forcatalyzing the initiation of protein synthesis by play-ing as a multipurpose ribosome adapter bridgingeIF4E (cap-binding protein), eIF3 (40S subunit bind-ing protein), eIF4A and Pab1p (poly-A bindingprotein), and an RNA unwinding helicase, respec-tively. Importantly, the abnormality in the proteinlevel or the activity of either initiation factor isknown to cause cell proliferation. We aim to test thepossibility of developing anti-eIF RNA aptamers fornovel diagnostic and therapeutic tools.

a. RNA aptamers to initiation factor 4A helicase hindercap-dependent translation by blocking ATP hydrolysis

The mammalian translation initiation factor 4A(eIF4A) is a prototype member of the DEAD-boxRNA helicase family that couples ATPase activity to

102

Uno, M., Ito, K., Nakamura, Y.: Polypeptide releaseat sense and noncognate stop codons by localizedcharge-exchange alterations in translational re-lease factors. Proc. Natl. Acad. Sci. USA, 99: 1819-1824, 2002.

Nakamura, Y., Uno, M., Toyoda, T., Fujiwara, T., Ito,K.: Protein tRNA mimicry in translation termina-tion. Cold Spring Harbor Symp. Quant. Biol., 66:469-475, 2002.

Kervestin, S., Garnier, O.A., Karamyshev, A.L., Ito,K., Nakamura, Y., Meyer, E., Jean-Jean, O.: Isola-tion and expression of two genes encoding eu-karyotic release factor 1 (eRF1) from Parameciumtetraurelia. J. Euk. Microbiol., 49: 374-382, 2002.

Nakamura, Y., Ito, K.: A tripeptide discriminator forstop codon recognition. FEBS Lett., 514: 30-33,2002.

Nevskaya, N., Tishchenko, S., Paveliev, M.,Smolinskaya, Y., Fedorov, R., Piendl, W.,Nakamura, Y., Toyoda, T., Garber, M., Nikonov,S.: Structure of ribosomal protein L1 fromMethanococcus thermolithotrophicus. Functionallyimportant structural invariants on L1 surface.Acta Cryst., D58: 1023-1029, 2002.

Ito, K., Fujiwara, T., Toyoda, T., Nakamura, Y.: Elon-gation factor G participates in ribosome disassem-bly by interacting with ribosome recycling factorat their tRNA-mimicry domains Mol. Cell, 9: 1263-1272, 2002.

Ito, K., Frolova, L., Seit-Nebi, A., Karamyshev, A.L.,Kisselev, L., Nakamura, Y.: Omnipotent decodingpotential resides in eukaryotic translation termi-nation factor eRF1 of variant-code organisms andis modulated by the interactions of amino acid se-quences within the domain 1. Proc. Natl. Acad.Sci. USA, 99: 8494-8499, 2002.

Blake, B.K., Ito, K., Nakamura, Y., Alam, S.: Back-bone 1H, 13C, and 15N assignments of the ribosomerecycling factor from Thermus thermophilus. J.Biomol. NMR, 24: 81-82, 2002.

Oguro, A., Ohtsu, T., Svitkin, Y., Sonenberg, N.,Nakamura, Y.: RNA aptamers to initiation factor4A helicase hinder cap-dependent translation byblocking ATP hydrolysis. RNA, In Press, 2003.

Polacek, N., Gomez, M., Ito, K., Xiong, L., Nakamura,Y., Mankin, A.: The critical role of the universallyconserved A2602 of 23S ribosomal RNA in the re-lease of the nascent peptide during translation ter-mination. Mol. Cell, 11: 103-112, 2003.

Nakamura, Y., Ito, K.: Making sense of mimic intranslation termination. Trends Biochem. Sci., 28:99-105, 2003.

Laursen, B.S., Siwanowicz, I., Larigauderie, G.,Hedegaard, J., Ito, K., Nakamura, Y., Kenney, J.,Mortensen, K.K., Sperling-Petersen, H.U.: Charac-terization of mutations in the GTP-binding do-main of IF2 resulting in cold-sensitive growth ofEscherichia coli. J. Mol. Biol., 326: 543-551, 2003.

和田美紀、中村義一：ニューモシスチス・カリニゲノムの可塑性。現代医療、34 (5): 1061-1065、2002．

中村義一：ゲノミクスとプロテオミクスを結ぶ RNA 情報ネットワーク（特集監修）。実験医学、20: 1382-1387,2002.

中屋敷徹、原英之、中村義一：酵母プリオンの分子生物学。医学のあゆみ、203: 938-943, 2002.

中村義一：RNA科学の潮流。「RNAの細胞生物学」（中村義一、坂本博編）、蛋白質・核酸・酵素、48: 319-328, 2003．

伊藤耕一、中村義一：翻訳終結の分子機構：いくつかのtRNA擬態狂騒曲への個人的視点。「RNAの細胞生物学」

（中村義一、坂本博編）、蛋白質・核酸・酵素、48: 329-337, 2003．

Publications

RNA binding and unwinding. In the crystal form,eIF4A has a distended ‘‘dumbbell’’ structure consist-ing of two domains, which probably undergo aconformational change, upon binding ATP, to form acompact, functional structure via the juxtaposition ofthe two domains. Moreover, additional conforma-tional changes between two domains may beinvolved in the ATPase and helicase activity ofeIF4A. However, the molecular basis of these confor-mational changes is not understood. Here, wegenerated RNA aptamers with high affinity foreIF4A by in vitro RNA selection-amplification. Uponbinding, the RNAs inhibit ATP hydrolysis. One classof RNAs contains members that exhibit dissociationconstant of 27 nM for eIF4A and severely inhibit cap-dependent in vitro translation. The binding affinity

was increased upon Arg substitution in the con-served motif Ia of eIF4A, which probably improves apredicted arginine network to bind RNA substrates.Selected RNAs, however, failed to bind either do-main of eIF4A that had been split at the linker site.These findings suggest that the selected RNAs inter-act cooperatively with both domains of eIF4A, eitherin the dumbbell or the compact form, and entrap itinto a dead-end conformation probably by blockingthe conformational change of eIF4A. The selectedRNAs, therefore, represent a new class of specific in-hibitors which are suitable for the analysis ofeukaryotic initiation, and which pose a potentialtherapeutic against malignancies that are caused byaberrant translational control.

103

Department of Basic Medical SciencesDivision of Molecular Biology (2)遺伝子動態分野 (2)

1. Restriction-modification gene complexes asselfish, mobile elements programming celldeath and genome rearrangement

A gene for a restriction (R) enzyme, which cutsDNA at a specific sequence, is often linked to a genefor a modification (M) enzyme, which methylates thesame sequence to protect it from cleavage. These sys-tems have been regarded as bacterial tools of defensethat attack invading unmethylated DNA but protectthe bacterium’s own methylated DNA. However,some RM gene complexes kill host bacteria that havethreatened their presence or lost them, using restric-tion cleavage of the chromosome. This and otherobservations led to the hypothesis that some RMgene complexes behave as selfish mobile genetic ele-ments, similar to viruses and transposons. Theincreasing evidence in support of this hypothesis in-cludes restriction site avoidance in bacteria, the lifecycle and mutual competition of RM complexes,their potential mobility and horizontal transfer, andtheir association with genome rearrangements.

a. Multiplication of a restriction modification gene com-plex

Marat Sadykov, Naofumi Handa, Yasuo Asami, Hi-

ronori Niki1, Masaru Tanokura2, Mitsuhiro Itaya3,and Ichizo Kobayashi: 1National Institute of Genet-ics, 2Department of Applied Biological Chemistry,Graduate School of Agricultural and Life Sciences,University of Tokyo, 3Mitsubishi Kagaku Instituteof Life Science

We found amplification of a restriction-modifica-tion gene complex. BamHI gene complex inserted intoBacillus subtilis chromosome showed resistance to re-placement by a homologous stretch of DNA. Somecells became transformed with the donor without los-ing BamHI RM. In most of these transformants,multiple copies of BamHI RM and the donor alleleswere arranged as tandem repeats. When a clone carry-ing one copy of each allele was propagated, extensiveamplification of BamHI RM and the donor units wasobserved only in the cells with a functional restrictionenzyme gene. This suggests restriction cutting of thegenome participates in the amplification. Visualiza-tion by fluorescent in situ hybridization revealed thatthe amplification occurred in single cells in a burst-like fashion that is reminiscent of induction ofprovirus replication. The multiplication ability in abacterium with a natural capacity for DNA release,uptake and transformation may contribute to spread-ing of RM gene complexes.

One genome is a community of genes potentially with different interests. Their collab-oration and conflicts underlie various life and death processes. Our goal is tounderstand genome dynamics, deaths, diseases, genome changes and genome evolu-tion from this point of view. Genes coding for a restriction-modification gene complexwill provide a clue to these problems. This understanding will open a door to a newtype of medicine and biotechnology.

Associate Professor Ichizo Kobayashi, Ph.D. 助教授薬学博士　小　林　一　三

104

b. Intragenomic movement of a restriction modificationgene complex

Seishi Ohashi, Noriko Takahashi-Kobayashi, YokoMizutani-Ui, and Ichizo Kobayashi

Potential mobility and horizontal transfer of restric-tion-modification gene complexes can be inferredfrom their linkage to mobile genetic elements, andfrom genome sequence comparisons and evolution-ary/informatics analyses. We detected movement of arestriction-modification gene complex within a cellwhen attempts were made to eliminate it from the cell.After blocking replication of a thermosensitive plas-mid carrying PaeR7I restriction-modification genecomplex in Escherichia coli by a temperature shift, ther-mo-resistant survivors carrying PaeR7I integrated intothe chromosome were recovered. The integration ap-peared to occur concomitantly with restriction attackon the chromosome and to be recA-dependent. Theproducts could be classified at the sequence level into:(i) those that had the plasmid and the chromosomeapparently co-integrated at chromosomal trans-posons (IS1 or IS5); (ii) those that were due to de novoinsertion of the IS1 together with the entire plasmidexcept for a 1-3 bp long deletion; (iii) those that result-ed from reciprocal crossing-over between the plasmidand the chromosome at a 1-3 bp region of homology.We discuss mechanisms of the underlying coopera-tion between restriction-modification genecomplexes, IS and host homologous recombinationfunctions. Thus restriction-modification gene com-plexes seem to use the same strategy of restrictionattack on the genome for their short-term and long-term persistence.

c. Molecular mechanism underlying post-segregationalhost cell killing by restriction modification enzymesystems

Asao Ichige and Ichizo Kobayashi

Previous work in this laboratory demonstrated thatcertain restriction modification gene systems can killhost cells when the gene systems are eliminated fromthe host cells. The molecular mechanism underlyingsuch post-segregational host killing by restrictionmodification gene systems is not completely under-stood, although several lines of evidence stronglysuggest that cleavage of the host chromosome by re-striction enzyme is responsible for the host cell killing.To further understand the molecular mechanism ofthe host cell killing by restriction modification genesystems, we are currently examining how cellular lev-el of restriction enzyme and that of modificationenzyme change when post-segragational host killingis induced.

d. Epigenetics in genome defense: a DNA methyltrans-ferase can protect the genome from post-disturbanceattack by a restriction-modification gene complex

Noriko Takahashi-Kobayashi, Yasuhiro Naito,Naofumi Handa, and Ichizo Kobayashi

Dcm in several bacteria methylates DNA to gener-ate 5’ CmCWGG. Vsr mismatch repair functionprevents C to T mutagenesis enhanced by this meth-ylation but promotes other types of mutation. Thereason for the existence of the dcm-vsr gene pair hasbeen unclear. We found that failure to replicateEcoRII restriction-modification genes, whose prod-ucts recognize the same sequence as Dcm, leads tochromosome degradation and loss of cell viability.This cell killing was suppressed by dcm. Dcm, there-fore, can play the role of a “molecular vaccine” bydefending the genome against parasitism by a re-striction-modification gene complex.

e. Death as a principle of symbiosis of genetic elementsin a genome — a hypothesis for involvement of mito-chondria in programmed cell death

Ichizo Kobayashi

Once some gene sets are established in a genome,their products kill the host organism when the per-sistence of these genes is threatened. A simpleexample is provided by a gene complex, such asEcoRI RM, encoding a restriction enzyme and a cog-nate modif icat ion methyl t ransferase . Thedescendants of cells that lose the RM gene complexare unable to modify a sufficient number of recogni-tion sites in their chromosomes to protect them fromlethal attack by the remaining molecules of restric-tion enzyme. When loss of a gene leads to death, thegene is called essential. However, since its first ap-pearance in the genome as a ‘dispensable’ gene, fromoutside or from inside, the gene may have co-evolved with the host so that it can now programdeath upon their curing. This form of programmedcell death, post-disturbance killing or addiction,might be a general principle in symbiosis of geneticelements within a genome. Eukaryotic programmedcell death often proceed through release of toxic mol-ecules from mitochondria. The capacity ofmitochondria to kill their host eukaryotic cell, upondisturbance, may have stabilized their symbiosis atthe initial stage. This scenario of symbiosis-through-death may provide a paradoxical answer to thequestion of how symbiosis can ever evolve from in-teraction between genetic elements with potentiallydifferent interests.

105

f. A nomenclature for restriction enzymes, DNA methyl-transferases, homing endonucleases and their genes

Richard J. Roberts4, Marlene Belfort5, TimothyBestor6, Ashok S. Bhagwat7, Thomas A. Bickle8,Jurate Bitinaite4, Robert M. Blumenthal9, SergeyKh. Degtyarev10, David T.F. Dryden11, Kevin Dyb-vig12, Keith Firman13, Elizaveta S. Gromova14,Richard I. Gumport15, Stephen E. Halford16, Stan-ley Hattman17, Joseph Heitman18, David P.Hornby19, Arvydas Janulaitis20, Albert Jeltsch21,Jytte Josephsen22, Antal Kiss23, Todd R. Klaenham-mer24, Ichizo Kobayashi, Huimin Kong4, Detlev H.Kruger25, Sanford Lacks26, Martin G. Marinus27,Michiko Miyahara28, Richard D. Morgan4, NoreenE. Murray29, Valakunja Nagaraja30, AndrzejPiekarowicz31, Alfred Pingoud32, Elisabeth Ra-leigh4, Desirazu N. Rao33, Norbert Reich34,Vladimir E. Repin35, Eric U. Selker36, Pang-ChuiShaw37, Daniel C. Stein38, Barry L. Stoddard39,Waclaw Szybalski40, Thomas A. Trautner41, JamesL. Van Etten42, Jorge M.B. Vitor43, Geoffrey G.Wilson4, Shuang-yong Xu4: 4New England Biolabs,5Molecular Genetics Program, New York StateDept. of Health, 6Genetics & Development, Colum-bia University, 7Department of Chemistry, WayneState University, 8Department of Microbiology,Biozentrum, Universitat Basel, 9Program in Bioin-formatics & Proteomics/Genomics, MedicalCollege of Ohio, 10SibEnzyme, 11School of Chemis-try, University of Edinburgh, 12Department ofGenetics, University of Alabama at Birmingham,13Biophysics Laboratories, School of Biological Sci-ences, University of Portsmouth, 14A.N. BelozerskyInstitute of Physico-Chemical Biology, MoscowState University, 15The University of Illinois Col-lege of Medicine, 16Department of Biochemistry,University of Bristol Medical School, 17Departmentof Biology, University of Rochester, 18HowardHughes Medical Institute, Duke University MedicalCenter, 19Department of Molecular Biology andBiotechnology, University of Sheffield, 20Institute ofBiotechnology, 21Institut fur Biochemie, Justus-Liebig-Universitat, 22Department of Dairy & FoodScience, Royal Veterinarian & Agricultural Uni-versity, 23Institute of Biochemistry, 24Departmentsof Food Science & Microbiology, 25North CarolinaState University, 26Institut fur Virologie-Charite,Humboldt Universitat, 27Brookhaven NationalLaboratory, 28Department of Pharmacology, Uni-versity of Massachusetts Medical School, 29NationalInstitute of Health Sciences, 30Institute of Cell andMolecular Biology, University of Edinburgh,31Centre for Genetic Engineering, Indian Instituteof Science, 32Institute of Microbiology,WarsawUniversity, 33Department of Microbiology and CellBiology, Indian Institute of Science, 34University ofCalifornia, Santa Barbara, 35State Research Centerof Virology & Biotechnology, 36Institute of Molecu-

lar Biology, University of Oregon, 37Department ofBiochemistry, The Chinese University of HongKong, 38Department of Microbiology, University ofMaryland, 39Fred Hutchinson Cancer ResearchCenter, 40McArdle Lab., University of Wisconsin,41MPI fur Molekulare Genetik, Ihnestrasse 42De-partment of Plant Pathology, University ofNebraska-Lincoln, 43Faculdade de Farmacia deLisboa

Since 1973, restriction endonucleases (Enases) andDNA methyltransferases (MTases) have been namedbased on an original suggestion of Smith & Nathans.They proposed that the enzyme names should beginwith a three-letter acronym in which the first letterwas the first letter of the genus from which the en-zyme was isolated and the next two letters were thefirst two letters of the species name. Extra letters ornumbers could be added to indicate individual strainsor serotypes. Thus, the enzyme HindII was one of fourenzymes isolated from Haemophilus influenzae sero-type d. The first three letters of the name wereitalicized. As more enzymes have been found, oftenfrom different genera and species with names whosethree-letter acronyms would be identical, consider-able laxity in naming conventions has appeared. Inaddition, we now know that each major type of en-zyme can contain sub-types. This especially applies tothe Type II enzymes, of which more than 3,500 havebeen characterized . In this work, we revisit the nam-ing conventions and outline an updated scheme thatincorporates current knowledge about the complexi-ties of these enzymes. A nomenclature will bepresented for restriction endonucleases, DNA methyl-transferases, homing endonucleases and related genesand gene products. It provides explicit categories forthe many different Type II enzymes now identifiedand provides a system for naming the putative genesfound by sequence analysis of microbial genomes.

2. Involvement of restriction-modification genecomplexes and other elements in genome re-arrangements and genome evolution assuggested from genome comparison

Recently complete sequences of two or more mi-crobial genomes that are closely related to each otherhave been determined. Detailed comparison of suchgenomes has become a useful approach for elucidat-ing principles and mechanisms of genome evolution.

a. CGAT: Comparative genome analysis tool for closelyrelated microbial genomes

Ikuo Uchiyama44, Toshio Higuchi45, MikihikoKawai, Yoko Mizutani-Ui, and Ichizo Kobayashi:44National Institute for Basic Biology, Okazaki Na-tional Research Institutes, 45INTEC Web andGenome Informatics Cooperation

106

For comparison of closely related bacterial ge-nomes, it is required not only to compare thesequences themselves but also to compare variousaspects of sequence features, combining with thealignment of the genomic sequences. To this goal, weare developing a genome analysis tool (CGAT). Inthis tool, an user can compare several feature seg-ments identified in each genome by varioussequence analysis programs, by overlaying themonto the alignment of homologous or orthologoussegments identified by all-against-all homologysearch.

b. Relation between restriction modification genes andgenome rearrangements suggested from genome se-quence comparison within genus Neisseria

Keiichirou Nakao, Akito Chinen, Ayaka Nobusato,Youhei Fujitani46, Ikuo Uchiyama44, and IchizoKobayashi: 46Department of Applied Physics &Physico-Informatics, Faculty of Science and Tech-nology, Keio University

The complete genome sequences of three differentlines in the genus Neisseria, became available – thoseof Neisseria meningitidis strain Z2491 (serogroup A),Neisseria meningitidis strain MC58 (serogroup B), anda Neisseria gonorrhoeae strain. We searched these ge-nomes for DNA methyltransferase homologues andcompared two genome sequences in their neighbor-hood by CGAT. Each genome was characterized byabundance of highly repetitive elements. In intraspe-cific comparison, we identified insertion of an RMgene complex into a putative operon. We also identi-fied insertion of a long DNA segment with an RMgene complex. In interspecific comparison, we iden-tified transposition of RM gene complex and moreexamples of operon insertion. In some cases, thepolymorphism is also linked with IS elements. Theseresults lend further support for the hypothesis thatRM genes are potentially mobile and involved in ge-nome rearrangements.

c. Mechanism of genome rearrangements and genomeevolution suggested from comparison between twocomplete genome sequences of Staphylococcus au-reus

Yoko Mizutani-Ui, Ikuo Uchiyama50, MikihikoKawai, Ichizo Kobayashi, Makoto Kuroda47, Toshi-ko Ohta48, Ikuo Uchiyama44, Tadashi Baba49,Harumi Yuzawa47, Ichizo Kobayashi, LongzhuCui47, Akio Oguchi49, Ken-ichi Aoki49, Yoshimi Na-gai49, JianQi Lian47, Teruyo Ito47, MutsumiKanamori48, Hiroyuki Matsumi Kanamori48, Hi-royuki Matsumaru48, Atsushi Maruyama48,Hiroyuki Murakami48, Akira Hosoyama49, YokoMizutani-Ui, Noriko Kobayashi, Toshihiro Tana-ka49, Toshihiro Sawano49, Ryu-ichi Inoue50,

Chikara Kaito50, Kazuhisa Sekimizu50, HidekiHirakawa51, Satoru Kuhara51, Susumu Goto52,Junko Yabuzaki52, Minoru Kanehisa52, Atsushi Ya-mashita53, Kenshiro Oshima53, Keiko Furuya53,Chie Yoshino53, Tadayoshi Shiba53, Masahira Hat-tori54, Naotake Osagasawa55, Hideo Hayashi56,Keiichi Hiramatsu47: 47Department of Bacteriolo-gy, Juntendo University, 48University of Tsukuba,College of Medical Technology and Nursing, 49Na-tional Institute of Technology and Evaluation,50Graduate School of Pharmaceutical Sciences,University of Tokyo, 51Faculty of Agriculture, Ky-ushu University, 52Institute for Chemical Research,Kyoto University, 53School of Science, Kitasato Uni-versity, 54Human Genome Research Group,RIKEN Genomic Sciences Center, 55Nara Instituteof Science and Technology, Graduate School of Bio-logical Sciences, Institute of Casic MedicalSciences, 56University of Tsukuba

Staphylococcus aureus is one of the major pathogenscausing both community-acquired and hospital-ac-quired infections. It produces a variety of toxins thatelicit both regional and systemic inflammation in hu-man body. Many of the toxins are known assuper-antigens that cause unique disease entitiessuch as toxic shock syndrome and staphylococcalscarlet fever. S. aureus has acquired resistance topractically all antibiotics so far introduced in clinicalpractice. Whole genome sequences of two relatedStaphylococcus aureus strains, N315 and Mu50, weredetermined. The two genome sequences were com-pared by CGAT to detect large-scale genomepolymorphisms. How they were generated was con-sidered. The restriction modification homologuesfound inserted into their drug-resistance island andinto their genomic islands were analyzed. A hypoth-esis for their evolution in relation to these mobileelements and for their role in toxin function and evo-lution was proposed.

d. Identification in a Methicillin-Susceptible Staphylo-coccus hominis of an Active Primordial MobileGenetic Element for the Staphylococcal CassetteChromosome mec (SCCmec) which – in Methicillin-Resistant Staphylococcus aureus – is the Carrier ofthe Resistance Gene mecA

Yuki Katayama47, Fumihiko Takeuchi57, TeruyoIto, Xiao Xue Ma47, Yoko Ui-Mizutani, IchizoKobayashi, and Keiichi Hiramatsu: 57Research In-stitute, International Medical Center of Japan andThe Organization for Pharmaceutical Safety andResearch

We previously reported that methicillin resistancegene mecA is carried by a novel type of mobile genet-ic element, SCCmec (Staphylococcal cassettechromosome mec) in the chromosome of methicillin-

107

resistant S. aureus (MRSA). These elements are pre-cisely excised from the chromosome and integratedinto a specific site on the recipient chromosome by apair of recombinases encoded by the ccr (cassettechromosome recombinase) A and B genes. In thepresent work, we detected homologues of the ccrgenes in S. hominis type strain GIFU12263 (equiva-lent to ATCC27844), which is susceptible tomethicillin. Sequence determination revealed thatthe ccr homologues in S. hominis were classified intotype-1 ccr genes (ccrA1, ccrB1) and were localized ona genetic element structurally very similar to SCCmecexcept for the absence of methicillin-resistance gene,mecA. The element had mosaic-like patterns of ho-mology with extant SCCmec elements, which wedesignated SCC12263 as a representative of type-1SCC. The ccrB1 gene identified in the S. hominis strainwas the first type-1 ccrB that retained its function asjudged from two criteria: 1) SCC12263 was spontane-ously excised during cultivation of the strain, and 2)introduction of the S. hominis ccrB1 into an MRSAstrain carrying type-I SCCmec whose ccrB1 gene isinactive generated SCCmec excisants at high frequen-cy. Existence of a SCC without mec determinantindicates a staphylococcal site-specific mobile genet-ic element that serves as a vehicle of transfer forvarious genetic markers across staphylococcal spe-cies.

3. Novel biotechnology based on behavior of re-striction-modification systems as selfishmobile elements

a. Application to maintenance and expression of usefulgenes

Noriko Takahashi-Kobayashi, Nanae Kotake58, Hi-roko Funaki58, and Masanori Watahiki58, IchizoKobayashi: 58Nippongene

The restriction modification (RM) gene pair has afunction to force their stable maintenance to theirhost. This provides the opportunity for stable main-tenance and expression of useful genes. Plasmidsthat carry lactose operon, a model useful gene clus-ter, connected to EcoRI RM genes (R+/R- and M+),were introduced into a lac- E. coli strain. The plasmidstability as well as LacZ activity were greatly in-creased by the presence of the R gene in the absenceof antibiotic selection. A similar stabilization in themaintenance and expression was observed withchloramphenicol acetyltransferase (CAT) gene at alarger industrial scale.

b. Experimental genome evolution

Youko Asakura59, Ichizo Kobayashi: 59Ajinomoto

Works from our laboratory demonstrated that an

attempt to replace chromosomally-located restric-tion-modification gene complex by a homologousstretch of DNA leads to a variety of large-scale ge-nome rearrangements. This may provide a novelprocedure of breeding of microorganisms.

c. Novel restriction enzymes from sequenced genomes

Ken Ishikawa, Masaru Tanokura60, Ichizo Koba-yashi, Toshihiro Kuroita61, Bunsei Kawakami61:60Department of Applied Biological Chemistry,Graduate School of Agricultural and Life Sciences,University of Tokyo, 61Toyobo

Comparison of closely related bacterial genome se-quences and other bioinformatic/ evolutionaryanalysis allowed us to predict putative restriction en-zyme genes in sequenced bacterial genomes. We areexpressing them in order to find out novel restrictionenzymes and to determine their three-dimensionalstructure.

4. Recombination machinery in the genomecommunity — homologous, site-specific andillegitimate — in interaction with restrictionmodification systems and other elements

Action of various machines of DNA recombina-tion in the cells is understood well in terms of conflictwith genetic elements within a genome such as re-striction modification systems.

a. Chromosomal site-specific recombination defendsgenome from post-segregational killing by a restric-tion-modification system

Yoji Nakamura, Asao Ichige, Naofumi Handa, andIchizo Kobayashi

XerCD forms a site-specific recombinase acting at aspecific site (dif) of E. coli chromosome. We found thatxerC and dif mutations enhance cell death after loss ofa restriction-modification gene complex. Our analysisof cell shape and chromosomes is in accord with thehypothesis that recombination repair of chromosomalrestriction breaks leads to a chromosomal multimer,which is resolved by the site-specific recombinationinto monomer chromosomes.

b. Double-strand-break-repair type homologous recom-bination by bacteriophages in gene manipulation invivo

Noriko Takahashi-Kobayashi, and Ichizo Koba-yashi

Earlier we demonstrated that a restriction break isrepaired by conservative (two-progeny) double-strand break repair in E. coli cells with active RecET

108

genes of Rac prophage or with active Red genes ofbacteriophage lambda. One recBC sbcB endA strain(BJ5183), which is popular in gene cloning frommammals and other eukaryotes, showed high activi-ty of conservative double-strand break repair inapparent contrast to our early demonstration of non-conservative (one-progeny) recombination in a recBCsbcB strain. We identified Rac prophage in this strainand demonstrated that mutational inactivation of itsrecT gene eliminates the double-strand break repairactivity. An endA mutation enhanced the double-strand break repair in a well established recBC sbcAstrain.

c. Alleviation of restriction by bacteriophage recombi-nation functions.

Naofumi Handa, and Ichizo Kobayashi

We now hypothesize that the conservative dou-ble-strand-break-repair recombination plays the roleto repair bacteriophage genomes after restriction at-tack. In support of the above hypothesis, recET genesof Rac prophage and red genes of bacteriophagelambda were shown to alleviate restriction.

d. Accumulation of large linear forms of bacterial chro-mosomes

Naofumi Handa, and Ichizo Kobayashi

Large linear forms of E. coli chromosome that areproduced by spontaneous breakage of its circularforms were detected by pulsed-field gel electro-phoresis in various recombination-related mutants.The results were interpreted in terms of the roles ofvarious proteins in the repair and processing of chro-mosomal breaks.

e. Non-homologous end-joining promoted by homolo-gous recombination function

Naofumi Handa, Ayumi Fujita-Kusano, Yoko Ui,Keiko Sakagami, and Ichizo Kobayashi

The above work and other works have revealed thata double-strand break is frequently generated and pro-cessed on the chromosomes from bacteria tovertebrates. The contrasting routes of its processing in-clude precise end-joining, non-homologousend-joining, and repair through homologous recombi-nation. We developed a sensitive assay fornon-homologous end-joining in bacterium, Escherichiacoli, and demonstrated presence of non-homologousend-joining promoted by homologous recombinationfunctions.

f. Homology-associated non-homologous recombina-tion

Kohji Kusano and Ichizo Kobayashi

We earlier identified non-homologous recombina-tion products that may have been generated bylong-range homologous interaction between twoDNAs in bacterial and mouse cells. We developed anassay systems in order to characterize this type of re-combination in E. coli. In one system, a plasmidcarrying inverted repeats, one with a type II restric-tion break and the other intact, was subjected to typeI restriction in vivo. Dependence on the rec genes wasdemonstrated, and the product structures were de-termined at the sequence level. The results providedsupport for the hypothesis of the illegitimate recom-bination dependent on homologous interaction.

5. Towards system biology of intra-genomicconflicts, genome dynamics and genome evo-lution

We employ mathematical approach, both analyti-cal methods and simulation methods, to understandessence of the above processes revealed by experi-mental work.

a. Asymmetric random-walk model in a reaction inter-mediate of homologous recombination

Youhei Fujitani46 and Ichizo Kobayashi

Homologous recombination can take place be-tween a pair of homologous regions of DNAduplexes. Its typical pathway begins with connectionof two strands coming from two recombining part-ners. A resultant connecting point (Hollidaystructure) migrates along the homologous region.We have formulated this migration in terms of a one-dimensional random-walk to succeed in explainingvarious phenomena: the dependence of recombina-tion frequency on the homology length, the mapexpansion, and the very rapid drop-off of recombi-nation frequency associated with sequencedivergence. Our model has supposed symmetric ran-dom-walk; its forward transition rate equals to itsbackward one. However, they can differ because ofpossible polarity of enzymatic machinery driving themigration. In this work, we took into account thisasymmetry in the random-walk model. Our analyti-cal results fitted well with the experimentalobservations with synthetic recombination interme-diates.

b. Random-walk model for interference in meiotic re-combination

Youhei Fujitani46, Shintaro Mori62, and IchizoKobayashi: 62Department of Physics, School of Sci-ence, Kitasato University

109

Kobayashi, I. Addiction in symbiosis of genetic ele-ments in a genome. In The Biology of Plasmids. B.Funnell and G. Phillips Ed. (American Society forMicrobiology Press), in preparation.

Kobayashi, I. Selfish behavior of restriction enzymes.In Restriction Enzymes. A. Pingoud (ed.) (SpringerVerlag), in preparation.

Kobayashi, I. Addiction as a principle of symbiosis ofgenetic elements in the genome — restriction en-zymes, chromosome and mitochondria. In Symbio-sis and Cellular Organelles. M. Sugiura, Ed. (Inter-national Society of Endocytobiology, Germany),in press.

Katayama, Y., Takeuchi, F., Ito T., Ma, X.X., Ui-Mizutani, Y., Kobayashi, I., Hiramatsu, K. Identifi-cation in a Methicillin-Susceptible Staphylococcushominis of an Active Primordial Mobile Genetic El-ement for the Staphylococcal Cassette Chromo-some mec (SCCmec) which – in Methicillin-Resis-tant Staphylococcus aureus – is the Carrier of theResistance Gene mecA. J. Bacteriol., in press.

Roberts, R. J., Belfort, M., Bestor, T. , Bhagwat, A. S.,Bickle, T. A., Bitinaite, J., Blumenthal, R. M.,Degtyarev, S. Kh., Dryden, D. T.F., Dybvig, K.,Firman, K., Gromova, E. S., Gumport, R. I.,

Publications

Halford, S. E., Hattman, S., Heitman, J., Hornby,D. P., Janulaitis, A., Jeltsch, A., Josephsen, J., Kiss,A., Klaenhammer, T. R., Kobayashi, I., Kong, H.,Kruger, D. H., Lacks, S., Marinus, M. G.,Miyahara, M. , Morgan, R. D., Murray, N. E.,Nagaraja, V., Piekarowicz, A., Pingoud, A., Ra-leigh, E., Rao, D. N., Reich, N., Repin, V. E., Selker,E. U., Shaw, P., Stein, D. C., Stoddard, B. L.,Szybalski, W., Trautner, T. A., Van Etten, J. L.,Vitor, J. M.B., Wilson, G. G., Xu, S. A nomenclaturefor restriction enzymes, DNA methyltransferases,homing endonucleases and their genes. NucleicAcids Res.in press.

Sadykov, M., Asami, Y, Niki, H., Handa, N., Itaya,M., Tanokura, M., Kobayashi, I. Multiplication of arestriction modification gene complex. Mol.Microbiol., in press.

Kobayashi, I. Life cycle of restriction-modificationsystems, powers in genome evolution. In GenomeScience — Towards a new paradigm? H.Yoshikawa, N. Ogasawara, N. Satoh, Eds.(Elsevier Science), 191- 200 (2003).

Takahashi, N., Yoshikura, H., Kobayashi ,I. An Es-cherichia coli strain, BJ5183, that shows highly effi-cient conservative (two-progeny) DNA double-

A crossing-over between homologous chromo-somes apparently suppresses another crossing-overin its neighborhood in meiosis. This crossover inter-ference or chiasma interference has been a subject ofvarious models — some physical/biological and oth-ers genetic/mathematical. We here propose a novelmodel that treats the process as a one dimensionalreaction-diffusion process. We suppose that an earlycontact point searches for global homology betweenhomologous chromosomes to initiate a crossing-over. We treat this contact point as a random-walkerthat moves along the homology, becomes immobi-lized and matures into a crossing-over point. Theinterference is caused by collision between the ran-dom-walker with another random-walker or with animmobilized point resulting in its destruction. Wenumerically showed that this simple model withonly two parameters — the initial density of the con-tact point per physical length, and the efficiency of itsprocessing into a crossing-over point — can describethe interference under a variety of conditions.

c. Co-evolution of bacterial restriction modification sys-tems and restriction sites on bacteriophage genomes

Akira Sasaki63, Ichizo Kobayashi, Ryota Horie64:63Department of Biology, Faculty of Science, Ky-ushu University, 64Riken

The genome decoding projects revealed that thegenomes are full of cis-elements such as transcrip-tion signals. Understanding their evolution is achallenging subject in genome biology and systembiology. The restriction sites along the genome mayprovide a simple system to study evolution of suchcis-elements in the genome sequence by selection.We constructed a mathematical model for popula-tion dynamics of bacteria carrying various restrictionmodification systems and bacteriophages carryingvarious restriction sites. We looked for conditions formaintenance of many restriction modification sys-tems in a bacterial genome and for conditions forevolution of recognitions sequences.

6. Basic studies for gene therapy by mutationcorrection

Asami Ino, and Ichizo Kobayashi

Earlier we demonstrated that adenovirus-mediat-ed gene transfer fol lowed by homologousrecombination with the genome can provide an effi-cient and accurate means of correcting mutations inmammalian cells. We have been trying to extend thisapproach to in vivo gene correction. We are also try-ing chimeric oligonucleotides for gene correction invivo. We are also analyzing rearrangements of viralvector genomes in mammalian cells.

110

strand break repair of restriction breaks. Gene, 303:89-97 (2003).

Fujitani, Y., Kobayashi , I. Asymmetric RandomWalk in a Reaction Intermediate of HomologousRecombination. J. Theor. Biol., 220: 359-370 (2003).

Takahashi, N., Naito, Y., Handa, N., Kobayashi, I. ADNA methyltransferase can protect the genomefrom post-disturbance attack by a restriction-modi-fication gene complex. J. Bacteriol., 184: 6100-6108(2002).

Fujitani, Y., Mori, S., Kobayashi , I. A reaction-diffu-sion model for interference in meiotic crossing-over. Genetics, 161: 365-372 (2002).

Kobayashi , I. DNA double-strand break repair inbacteria. In the Encyclopedia of Life Sciences,(Macmillan Reference Limited, London), (2002).

小林一三：制限酵素修飾酵素遺伝子の自己増殖の発見。海海海海海洋洋洋洋洋、刊行中。

小林一三：ゲノムにとって「自己」とは何か？　ゲノムゲノムゲノムゲノムゲノム・・・・・ニュ－スニュ－スニュ－スニュ－スニュ－ス、刊行中。

小林一三、広瀬彩：バクテリオファージ・ラムダ。細胞生物学事典、朝倉書店、刊行中。

小林一三：メイナードスミス博士京都賞受賞記念ワークショップ講演：ゲームの場としてのゲノム。　遺伝遺伝遺伝遺伝遺伝、56,70-76 (2002).

111

Department of Basic Medical SciencesDivision of Molecular Biology (3)遺伝子動態分野 (3)

1. Proteolysis and cell death

Junko Ohmoto and Shinobu Imajoh-Ohmi

Cell death involves various intracellular pro-teolytic enzymes such as caspase, a series of cysteineproteases cleaving substrates after aspartate residue;proteasomes, a protein hydrolysis system regulatedby ATP and ubiquitin; and calpain, calcium-depen-dent protease existing in the cytosol as an inactiveprecursor form. Among them caspases are now es-tablished as pivotal apoptosis-executing enzymesthat cleave various substrates. Endogenous or viralproteins and synthetic substances inhibitory forcaspases suppress the apoptotic cascade and rescuecells from cell death. On the other hand, proteasomesdrive the cell cycle by degrading cyclins etc., and alsoplay important parts in apoptosis, since proteasomeinhibitors induce apoptotic cell death in growingcells but suppress apoptosis of some cells that is inquiescent state. Furthermore, in some specific cellssuch as polymorphonuclear leukocytes, other pro-teases might be involved in cell death.

Polymorphonuclear neutrophils (PMNs) undergospontaneous apoptosis during cultivation in vitro.Various proteases are also activated and many target

proteins have been reported in apoptotic PMNs. Ac-tin is proteolyzed to a 40-kDa fragment that lacksamino-terminal region involved in polymerization.To investigate the role of actin proteolysis we made acleavage-site-directed antibody for the 40-kDa formof actin using synthetic peptide as a hapten. The anti-body stained the 40-kDa polypeptide but did notrecognize native actin abundant in cell lysates. First,we found that the 40-kDa fragment is generated dur-ing isolation of PMNs from peripheral bood. Byusing diisopropyl fluorophosphate, an inhibitor forserine proteases, PMNs with native actin could beprepared. Furthermore, elastase was identifited asthe enzyme responsible for the limited proteolysis ofactin. In fact, when isolated PMNs were incubatedwith elastase, the 40-kDa fragment was observed,providing us with a question how extracellularelastase attacks actin.

2. Phagocytic differentiation and apoptosis

Yuichi Niikura and Shinobu Imajoh-Ohmi

Among phagocytes macrophages are long-lived toplay important roles in the defense system of the hostafter settlement in various tissues. On the other

Associate Professor Shinobu Imajoh-Ohmi, D. Sc. 助教授理学博士　大　海　　　忍

To understand various cellular phenomena on the basis of structure and function ofproteins, we have developed novel antibodies that discriminate post-translationalmodification of proteins such as phosphorylation and limited proteolysis. The powerfulimmunocytochemical probes visualize emzymatic reactions in situ and enable us toperform biochemical analysis of growing, differentiating and dying cells without anycell sorting.

112

hand, polymorphonuclear leukocytes spontaneouslyundergo apoptosis. Using a human monoblast U937cell line we have investigated relationship betweendifferentiation and cell death. The U937 cells differ-entiate to superoxide anion-producible cells aftercultured with interferon γ (IFN), 1α, 25-dihydroxyvitamin D3 (VD3) or retinoic acid (RA). RA- and VD3-differentiated U937 cells seem to be more closelyrelated to mature macrophages than IFN-treatedcells, since RA and VD3 induce cell surface expres-sion of CD11b. IFN-treated U937 cells become highlysensitive to Fas-mediated apoptosis. On the otherhand, RA- and VD3-differentiated cells showed resis-tance against cytotoxic anti-Fas antibody. In thesecells, the appearance of caspase activation coincideswith Fas susceptibility. Here we addressed whethercaspase activation through the mitochondrial cas-cade is functionally active in differentiated U937cells. Within the cytosolic fractions of the differenti-ated cells, caspase-3 and caspase-7 were fullyactivated in response to cytochrome c and dATP. Im-portantly, the processing of caspase-8 followed bythe cleavage of BID was also observed in these sam-ples, possibly allowing the amplification of caspaseactivation by re-stimulation of the mitochondrialpathway. These results demonstrate that once cyto-chrome c is released to the cytosol, caspases areactivated equally in differentiated cells, irrespectiveof their Fas susceptibility. In mitochondria isolatedfrom differentiated cells, the addition of activecaspase-8 in the presence of the cytosolic fraction in-duced mitochondrial membrane depolarizationfollowed by cytochrome c release. This demonstratesthat the mitochondria from differentiated cells sus-tain the ability to carry out cell death. In addition,staurosporine induced caspase activation in VD3-and RA-treated cells, demonstrating that the mito-chondrial pathway is functional in vivo as well as invitro. Fas expression was down-regulated in VD3-and RA-treated cells in contrast to IFN-treated cellsin which Fas expression was up-regulated. Consis-tent with these data, Fas clustering induced by theagonistic anti-Fas antibody CH-11 was extremely re-duced in VD3- and RA-treated cells. We concludethat the amplification loop for caspase activationthrough mitochondria is functionally active in differ-entiated U937 cells, and suggest that changes in Fassusceptibility occur around the initial steps of deathsignaling upstream of the mitochondrial pathway.

3. Cell death of monocytic cells and macrophag-es infected with Shigella flexneri

Upon invasion into tissues bacterial phathogensare phagocytosed by resident macrophages to bekilled and digested. Some bacteria can escape intothe cytosol and induce cell death of the host cell. Shi-gella flexneri is also reported to induce apoptosis in

murine macrophages where a bacterial invasionplasmid antigen B (IpaB) activates a cellular proteasetriggering apoptotic cell death. However, other in-vestigators observed necrotic cell death inShigella-infected human macrophages derived fromperipheral monocytes. Cell death of macrophagescaused by bacterial invasion remains to be character-ized on the basis of molecular interaction.

a. Internalization of virulent Shigella is required for in-duction of rapid necrosis to macrophage-like cells,and apoptosis is induced by extracellular bacteria in-dependently on Shigella’s pathogenicity

Takashi Nonaka, Hitomi Mimuro1, Asaomi Ku-wae1, Toshihiko Suzuki1, Chihiro Sasakawa1 andShinobu Imajoh-Ohmi: 1Division of Bacterial In-fection, Department of Microbiology andImmunology

It is currently unclear whether Shigella kills its hostcells by apoptosis or necrosis. Here we show that rap-id necrosis ensues in macrophage-like cell lines (U937cells differentiated by all trans-retinoic acid and J774cells) infected with the S. flexneri strain YSH6000. Theinfected cells rapidly lose membrane integrity, a typi-cal feature of necrosis, as indicated by the release ofthe cytoplasmic lactate dehydrogenase and the expo-sure of phosphatidylserine (PS) associated with therapid uptake of propidium iodide (PI). The infectedcells exhibit DNA fragmentation without nuclear con-densation and substantial involvement of eithercaspase-3/-7 or caspase-1 was not detected, which isalso contrary what is normally observed in apoptosis.Cytochalasin D potently inhibited Shigella-inducedcell death, indicating that only internalized Shigellacan cause necrosis. Osmoprotectants such as polyeth-ylene glycols could suppress cell death, suggestingthat insertion of a pore by Shigella into the host cellmembrane induces the necrosis. The pore was esti-mated to be 2.87±0.4 nm in diameter. Shigella was alsofound to be able to induce apoptosis but only in one ofthe lines tested and under specific conditions, namely,U937 cells differentiated with interferon-γ (U937IFN).Caspase-3/-7 but not caspase-1 activation was ob-served in these infected cells and the exposure of PSoccurred without the uptake of PI. An avirulent Shi-gella strain, wild-type Shigella killed with gentamicin,and even E. coli strain JM109 could also induce apop-tosis U937IFN cells, and cytochalasin D could notprevent apoptosis. It appears therefore that Shigella-induced U937IFN cell apoptosis is unrelated toShigella pathogenicity and does not require bacterialinternalization. Thus, Shigella can induce rapid necro-sis macrophage-like cells in a virulence-relatedmanner by forming pores into host cell membranewhile some cells can be killed through apoptosis in avirulence-independent fashion.

113

b. Toll-like receptor 4 (TLR4) is up-regulated in IFN-t reated U937 cel ls and possibly mediateslipopolysaccharide-induced apoptotic cell death

Taku Kuwabara and Shinobu Imajoh-Ohmi

In Shigella-infected monocytes/macrophages two-types of cell death occur and compete eath other; oneis apoptosis and other non-apoptotic cell death trig-gered by pore formation resulting in disruption ofmenbrane function. Precultured with IFN and there-after infected with Shigella , the U937 cellspredominantly exhibit apoptosis that is independentof virulence of the pathogen. On the other hand, celldeath depends on virulence in RA- or VD3-differenti-ated cells where apoptosis is suppressed. We havefound here a bacterial component lipopolysaccharide(LPS) could induce apoptosis in IFN-preculturedU937 cells. Cell surface expression of the LPS receptor,Toll-like receptor 4 (TLR4), augmented in IFN-treatedcells but not in RA- or VD3-treated ones. Antagonisticantibody against TLR4 rescued cells from LPS-in-duced apoptosis but not from cytotoxic anti-Fas.Furthermore, treatment of the cells with antisense oli-gonucleotide for TLR4 decreased sensitivity to LPS inparallel with decreasing expression of the LPS recep-tor. These findings strongly suggest that sensitizatonto LPS-mediated apoptosis is due to up-regulation ofTLR4 expression in IFN-cultured U937 cells. Apoptot-ic signalling pathway from the LPS receptor is underinvestigation.

4. Identification and characterization of a novelinhibitor of topoisomerase IIααααα (ITIIααααα)

Akira Nakanishi2,3, Fumio Hanaoka2,3,4 and Shino-bu Imajoh-Ohmi: 2CREST, Japan Sci. andTech.Corp., 3Cell. Phys. Lab., RIKEN, 4Grad. Sch.Front. Biosci., Osaka Univ.

DNA topoisomerase IIα is essential for properchromosome condensation during mitosis and forsegregation of sister chromatids during anaphase.This activity alters DNA topology during catena-tion/decatenation, enables knotting/unknotting,and relaxes DNA supercoils generated during DNAreplication and RNA transcription. TopoisomeraseIIα is expressed in a cell cycle-dependent manner.Protein levels are greater in proliferating cells than inquiescent cells, and are lowest during G1. Levels be-gin to increase prior to S phase, remain relativelystable through S, and increase again and peak in lateG2.

Previous studies have shown that GyrI, a regula-tory factor of DNA gyrase activity, inhibits thesupercoiling activity of DNA gyrase (prokaryotictype II topoisomerases) and that both overexpressionand antisense expression of the gyrI gene suppresscell proliferation. We have identified a cDNA encod-

ing a novel inhibitor of topoisomerase IIα (ITIIα),which is homologous to E. coli DNA gyrase inhibitor(GyrI). Baculovirus expressed recombinant ITIIαprotein binds and inhibits the activity of topoi-somerase IIα. Using antibodies specific for ITIIα andtopoisomerase IIα, we used immunohistochemistryto show that endogenous ITIIα in normal human fi-broblast cells accumulated in the cytoplasm fromG0/G1 to S-phase, and translocated to the nucleus atG2/M, whereas topoisomerase IIα was localized tothe nucleus at G2/M. Furthermore, we have foundreduced expression of ITIIα in many types of carci-nomas (HeLa, Lung, Colon, Gallbladder, Rectal,Renal) compared with normal human fibroblastcells, and that down-regulation of ITIIα followingaddition of ITIIα antisense oligonucleotide correlateswith increased expression of topoisomerase IIα pro-tein. However, the levels of topoisomerase IIαmRNA were unchanged following reduction of ITIIαlevels by antisense treatment, indicating that the ef-fects on the levels of topoisomerase IIα were not dueto an effect on its transcription or mRNA stability.These findings suggest that ITIIα- topoisomerase IIαinteraction may provide a molecular basis for key ac-tivities of topoisomerase IIα.

5. Establishment of novel antibodies as toolsavailable for in situ analyses of post-transla-tional modification of proteins

After biosynthesis proteins undergo various post-translational modifications, and their functions aremodulated. In order to understand such biochemicalreactions in a single cell, we have been making mod-ification-specific antibodies as probes for in situanalyses; cleavage-site-directed antibodies for pro-teolysis, phosphorylation-site-specfic antibodies,myristoilated peptide-specific antibodies, ubiquiti-nation-specific antibodies, inhibitor-boundenzyme-specific antibodies etc. These antibodiesshould be useful tools for research in cellular bio-chemistry.

a. Cleavage-site-directed antibodies

Yuichi Niikura, Takashi Nonaka, and ShinobuImajoh-Ohmi

We have previously demonstrated that technics ofpeptide synthesis and anti-peptide antibody produc-tion enable us to obtain antibodies to neoantigensgenerated by proteolysis. Such cleavage-site-direct-ed antibodies specifically bind to terminal regions ofproteolyzed fragments including either amino orcarboxyl group newly ionized by hydrolysis of thepeptide bond. The most remarkable characteristic ofcleavage-site-directed antibodies is that they do notcross-react with unproteolyzed native polypeptidesalthough the same sequence exists internally in the

114

polypeptide. The strict specificity of antibodies ga-rantee in situ analysis of proteolysis withoutfractionation of proteins by biochemical methods. Toobtain a cleavage-site-directed antibody a syntheticpeptide mimicking the terminal region of a proteo-lyzed protein is used as a hapten, where moleculardesign of the hapten is critical. We have so far estab-lished cleavage-site-directed antibodies for variousproteins: active forms of calpains, calcium-depen-dent proteases with high- and low-calciumsensitivities; calpain-catalyzed fragments of proteinkinase C species; compliment component C1s andcaspase-catalyzed poly(ADP-ribose) polymerase inapoptotic cells.

Death receptor-mediated apoptosis involves pro-teolytic activation of procaspase-8 via interaction ofadaptor protein FADD with the two proteins. FLICE-like inhibitory protein (FLIP) was first identified asan inhibitor for caspase-8, since it lacks an SH groupin the active center. FLIP is also a substrate forcaspase-8. We generated and characterized novel an-tibodies specific for a cleavage site of humancaspase-8/FLICE and its substrate FLICE-like inhibi-tory protein (FLIP). The synthetic peptides used asimmunogens were CQGDNYQKGIPVETD (#791)and VSEGQLEDSSLLEVD (#1342), which corre-sponded to a cleaved region of an N-terminalfragment of caspase-8 and FLIP generated by activecaspase-8, respectively. Each antibody purified fromrabbit antiserum reacted specifically with the immu-nogen but not with the peptide corresponding to theunproteolyzed form, as assessed by ELISA. In vitrocleavage of GST-FLIP by active caspase-8 generatedan N-terminal (GST-p43) and a C-terminal fragment(p12). Consistent with other in vivo data, the FLIPc l e a v a g e s i t e f o l l o w s t h e A s p r e s i d u e ,LEVD376GPAMKNVEF, identified by N-terminalsequencing of the p12 fragment. #1342-antibody(#1342-Ab) recognized the GST-p43 fragment but notthe uncleaved protein, thus confirming its specifici-t y . W h e n t h e a n t i b o d i e s w e r e u s e d i nimmunoblotting, flow cytometry, and confocal lasermicroscopy, the proteolysis of caspase-8 and FLIPand the subcellular localization of their digests couldbe monitored in apoptotic U937 cells. Interestingly, asignificant rise in the percentage of cells exhibitingcaspase-8 and FLIP cleavages was observed uponFas stimulus in interferon-gamma-treated U937 cells,in which the susceptibility to Fas is extremely en-hanced. In contrast, U937 cells treated with vitaminD3 or all-trans retinoic acid showed Fas-resistanceand caspse-8 processing and FLIP cleavage werestrongly inhibited. In conclusion, we established asystem based on the cleavage site-directed antibod-ies to monitor the dynamics of caspase-8 processingand activation during apoptosis. Using the systemwe found that Fas-susceptibility changes duringU937 differentiation occur upstream of caspase-8processing/activation.

b. A novel method for hunting substrates of limited pro-teolysis

Masahiko Kato, Hiroyuki Fukuda, Takashi Nona-ka and Shinobu Imajoh-Ohmi

During the course of study on calpain/calpastatinsystem in apoptosis we have found that a cleavage-site-directed antibody recognizes a novel moleculeunrelated to the expected target protein. To analyzeintracellular mobilization of calpastatin antibodieswere raised against peptidyl haptens mimicing ter-minal regions of calpastatin polypeptides generatedby caspases. A cleavage-site-directed antibodystained the amino-terminal 30-kDa fragment of re-combinant human calpastatin cleaved in vitro bycaspase-7. However, calpastatin was not detected bythe same antibody in apoptotic cells, suggesting thatthe calpastatin fragment underwent further degra-dation. Instead, a 95-kDa polypepted wasrecognized by immunoblotting with this antibodyduring apoptosis. The 95-kDa band was seen specifi-cally in apoptotic cells, and diminished in thepresence of caspase inhibitors. Under less stringentconditions a 110-kDa polypeptide was also observedin non-apoptotic cells, but decreased in apoptoticcells in parallel with appearance of the 95-kDa band,suggesting that the 110 kDa protein was cleaved to95K by caspases during apoptosis. By further struc-tural analysis of the two antibody-stainedpolypeptides by Edman degradation and mass spec-trometry, however, the 110 kDa and 95-kDapolypeptides were identified as APG-2, a member ofheat shock protein, and a caspase-cleaved heavychain of myosin II-A, a non-muscle type myosin, re-spectively. Furthermore, we found several targets forcaspases, some of which remain to be identified, byanother type of cleavage-site-directed antibodies.

c. Proteomic approach for identification of cysteine pro-teases in Caenorhabditis elegans

Jin Ling, Hiroyuki Fukuda and Shinobu Imajoh-Ohmi

E64c, [L-3-trans-carbonyloxirane-2-carbonyl]-L-leu-cine(3-methylbutyl)amide, is a synthetic inhibitor forcysteine proteases such as cathepsins B, H, L andcalpain. To inhibit intracellular cysteine proteasesE64d, [L-3-trans-ethoxycarbonyloxirane-2-carbonyl]-L-leucine(3-methylbutyl)amide, a membrane-permeablederivative of E64c is used instead of E64c. E64d pene-trates into the cell where cellular esterases convert it toE64c that covalently binds to the SH group of activecenter in enzymes. Thus, anti-E64c antibody is a usefulprobe for in vivo analysis of cysteine proteases.

We have succeeded in making an antibody toE64c. First, we tried to establish an antibody againstE64c-bound calpain. A peptide corresponding to the

115

active center of calpain was synthesized by using themultiple-antigen peptide system. E64c was chemi-cally introduced into the SH group of active centercysteine under reducing conditions. Rabbits wereimmunized with the E64c-conjugated calpain-de-rived peptide without further conjugation with acarrier protein. Unexpectedly, an antibody thus pre-pared reacted not only with E64c-inactivated calpainbut also with E64c-bound other cysteine proteasessuch as papain and cathepsins. Low antigenicity ofpeptide region in the immunogen may result in suchbroad specificity of the antibody. Our antibody is ex-pected to be used for identification of E64c-targetednovel proteases. When cells were treated with E64d,

cell growth was suppressed and several proteinswere labeled by E64c that is visualized with this anti-body on immunoblotting. Structural analysis ofthese proteins may lead identification of novel cys-teine proteases.

Homogenates of C. elegans were treated with E64cin the presence or absence of calcium ion, and sub-jected to electrophoresis/immunoblotting using ananti-E64c antibody. A 55-kDa polypeptide (p55) waslabelled with E64c in a calcium ion-dependent man-ner. In C. elegans several calpain-related geneproducts were identified at the mRNA level, buttheir physiological function remains to be elucidat-ed. p55 is to be analyzed by mass spectrometry.

Yamamoto K, Inoue, N Masuda, R Fujimori A, SaitoT, Imajoh-Ohmi S, Shinkai H and Sakiyama H.Cloning of Hamster Type XVII Collagen cDNA,and Pathogenesis of Anti- Type XVII CollagenAntibody and Complement in Hamster BullousPemphigoid. 2002. J Invest Dermatol 118: 485-492.

Tanaka Y, Ishikawa F, Osada H, Imajoh-Ohmi S,Uchida T and Kakiuchi T. Reveromycin A inhibitsantigen receptor-mediated antigen presentationby B lymphoma cells via its effect on intracellulartrafficking of the antigen. 2002. J Antibiot 55: 904-913

Sakiyama H, Nonaka T, Masuda R, Inoue N, KubokiY, Iijima M, Hirabayasi Y, Takahagi M, Yoshida K,Kuriiwa K, Yoshida M and Imajoh-Ohmi S. Char-acterization of mineral deposits formed in culturesof a hamster tartrate-resistant acid phosphatase(TRAP) and alkaline phosphatase (ALP) double-positive cell line (CCP). 2002. Cell Tissue Res 309:269-279.

Publications

Niikura Y, Nonaka T and Imajoh-Ohmi S. Monitor-ing of Caspase-8/FLICE Processing and Activa-tion upon Fas Stimulation with Novel AntibodiesDirected against a Cleavage Site for Caspase-8 andIts Substrate, FLICE-Like Inhibitory Protein(FLIP). 2002. J Biochem 132: 53-62.

Nakanishi A, Imajoh-Ohmi S and Hanaoka F. Char-acterization of the Interaction between DNAGyrase Inhibitor and DNA Gyrase of Escherichiacoli. 2002. J Biol Chem 277: 8949-8954.

Azumi H, Inoue N, Ohashi Y, Terashima M, Mori T,Fujita H, Awano K, Kobayashi K, Maeda K, HataK, Shinke T, Kobayashi S, Hirata K, Kawashima S,Itabe H, Hayashi Y, Imajoh-Ohmi S, Itoh H andYokoyama M. Superoxide generation in direc-tional coronary atherectomy specimens of patientswith angina pectoris: important role of NAD(P)Hoxidase. 2002. Arterioscler Thromb Vasc Biol 22:1838-1844

116


1. Ex vivo expansion of haematopoietic stemcells using human GM-CSF receptor Fall mu-tant transgenic mouse

Yutaka Aoki1, Ken-ichi Arai1, Sumiko Watanabe1:1Department of Clinical Oncology

Using Tg mice expressing hGM-CSFR (wild-Tg),we previously showed that hGM-CSF is not a GM-lineage promoting factor but a strong proliferationpromoting factor of all the myeloid lineages exam-ined. To analyse signaling requirement for theseactivities, we generated Tg mice expressing FallhGM-CSF mutant receptor, in which all the cytoplas-mic tyrosine residues of hGM-CSF receptor β subunitwere replaced by phenylalanine. Methyl cellulose as-says of bone marrow cells from Fall-Tg, wild-Tg andtheir litter mate showed that Fall induced lower butstill significant numbers of myeloid colonies. Inter-estingly, immature colonies were major populationof Fall induced colonies and differentiated coloniessuch as erythroid colonies were not observed. Meth-yl cellulose assay of 5-FU treated mouse furtherconfirmed this tendency. We isolated lin- bone mar-

row cells of Fall-Tg as well as wild-Tg cells and cul-tured these cells in the presence of hGM-CSF andSCF. After 1 week of culture, significantly large num-er of scaI+,c-kit+ cells were observed with the Fall-Tgderived bone marrow cells. Taken together, it wassuggested that the Fall signals can be used to expandhaematopoietic stem cells by ex vivo culture.

2. Cloning and functional Analyses on a NovelPH Domain Protein

Eishun Muto, Ken-ichi Arai, Sumiko Watanabe

To study the development of nervous system, dif-ferential expression cloning based on degenerateRT-PCR was done. One of the cloned genes, Clone22,was identified as a norvel PH domain protein. Ex-pression analyses revealed that its expression ishighly restricted in eye and kidney in adult mice, andthat, during nervous system development, high-lev-el expression is observed in ventricular zone of brainand spinal cord.

To address the function of Clone22, Loss of func-tion analysis by antisense MO oligo in zebrafish was

Department of Molecular andDevelopmental Biology染色体制御分野

Our long-term goal is to understand the molecular mechanisms which coordinatelyregulate growth and differentiation of stem cells as well as differentiated cells withemphasis on intracellular signal transduction. For this purpose, we are using systemsranging from zebrafish, chick, mouse and culture cells. The major research areas ofinterest are on: 1) development and regeneration of eye, 2) roles of cytokines and theirreceptors in hematopoietic stem cells, 3) Th1 and Th2-specific cytokine genes in acti-vated T cells, 4) regulation of DNA replication and cell cycle. On the basis of theseefforts, we intend to develop technologies to manipulate growth and differentiation ofvarious stem cells with high fidelity, which is important for cell and gene therapy.

Professor　 Ken-ichi Arai, M.D., Ph.D.Associate Professor Sumiko Watanabe, Ph.D.Research Associate Noriko Sato, M.D., Ph.D.Research Associate Shinya Satoh, Ph.D.

教　授医学博士新　井　賢　一助教授医学博士　渡　辺　すみ子助　手医学博士　佐　藤　憲　子助　手工学博士　佐　藤　伸　哉

117

done. MO-injected embryo showed expanded area offorebrain suggested the possibility that clone 22 wasrequired for the proper development of neural sys-tem. To further character ize i ts funct ion,construction of gene-targeted deletion mouse (KOmouse) have been performed.

3. Analysis of genes of which expression is reg-ulated in developing mouse eye

Akihiko Muto, Rika Saito, Ken-ichi Arai and Sum-iko Watanabe

To analyze eye development, we did differentialdisplay and identified two genes, DD29 and DD76,of which expression level is varied during mouse eyedevelopment. Mouse DD29 (mDD29) was predictedas a serine/threonine kinase with a putative kinasedomain and a leucine-zipper motif at N- and C-ter-mini, respectively. While mDD29 was expressed inthe subventricular zone of telencephalon and inwhole neural retina during embryonic period (E12.5to E17.5), the expression was sharply decreased afterbirth. mDD29 expression was also detected in multi-ple other tissues of postnatal mice. Mouse DD76(mDD76) consists of 635 amino acids and containsone EF-hand and two leucine-zipper motifs withinN- and C-terminal regions, respectively. Tissue ex-pression in P1 and adult mice examined by RT-PCRdemonstrated that mDD76 was exclusively ex-pressed in neural tissues including brain and eye.Analysis of temporal expression profile in eyes by insitu hybridization indicated that mDD76 transcriptwas first detected in retinal pigment epithelium atE10.5, and then in neural retina at later stages. FromE13.5 to P1, mDD76 was expressed through entireretina, and, with the progression of development, theexpression was gradually intensified in an inner sideof retina where ganglion and amacrine cells are lo-cated. In adult mouse retina, mDD76 mRNA wasreduced to an undetectable level.

Isolation and analyses of zebrafish orthologues(zDD29 and zDD76) manifested that the primarystructure and expression pattern of each gene werewell conserved in zebrafish. To examine the func-tional role of these genes in development, weperformed the knockdown experiment by usingmorpholino antisense oligo (MO). Injection ofzDD29-specific morpholino (zDD29-MO) in ze-brafish embryo gave rise to abnormal swellingaround tectum region at 24 hours post fertilization(hpf). Defects in the blood circulation system includ-ing delayed onset of blood circulation, decreasednumber of blood cells, and pericardiac edema werealso observed. These results indicated that DD29might function in brain and blood cell development.On the other hand, knockdown of zDD76 by zDD76-MO resulted in small sizes of head and eyes around30 hpf, whereas the size and morphology of trunk

were essentially normal. Concomitantly, defects inblood circulation and in morphology of tail were ap-peared in a subset of MO-injected embryos. Detailedanalysis of the eye development revealed that retinallamination was disorganized and differentiation ofphotoreceptor cells was significantly retarded byMO injection, suggesting that DD76 plays importantroles in retinal development.

4. Roles of p53-related gene, p73 in zebrafishdevelopment

Shinya Satoh, Ken-ichi Arai, Sumiko Watanabe

p73 is one of the p53-related genes and has beensuggested to be important for neurogenesis in mousedevelopmental process. Multiple splicing forms ofp73 had been reported but molecular mechanismsand signaling pathways of these variants in p73functions has not been well documented. To califydetailed mechanism of p73 in vertebrate develop-mental process, we took advantages of zebrafishsystem. We isolated two alternative splicing forms ofzebrafish p73 by RT-PCR and found that one isoformcorresponded to α form which was reported in hu-man and mouse. In addition to the α form which ismajor product of p73, we found a new splicing vari-ant, which produces C-terminal deleted form ofp73α protein. We analyzed transcriptional activationability of these zp73 isoforms using p53-responsiveelement reporter gene in mammalian cells. Interest-ingly, we found that the levels of transcriptionalactivity of the isoforms are different. Currently, weare anlyzing the detailed expression patterns of p73variants in developing zebrafish embryo, and startedto do loss-of function experiments using MO-anti-sense oligo. By using Mo designed to variousdifferent position of zp73, we aimed to knockdowncertain specific form of zp73 isoforms.

5. Isolation and functional analysis of a novelgene related to retinal development

Ryo Kurita, Hiroshi Sagara2, Yutaka Aoki, Ken-ichi Arai and Sumiko Watanabe: 2Department offine morphology

In order to understand the molecular mechanismsof retinal development, we planned to isolate novelgenes which are specifically expressed in the retina.We randomly picked up unknown sequences fromzebrafish adult retina EST database and examinedexpression pattern of these clones by RT-PCR to se-lect genes with head-specific expression. Theseclones were further characterized by RACE and insitu hybridization. Among these clones, we focusedon #61, which encodes a novel actin-binding protein.During the zebrafish embryogenesis, #61 expressionwas first detected in lateral mesoderm of the mid-

118

trunk region, and then, it became to be restricted intosome parts of brain and choroid fissure of the eyealong with embryogenesis. Functional knock-downof #61 by using an antisense morpholino oligo (#61-MO) resulted in morphological defects of choroidfissure formation as well as in curly thickened tailsand yolk sac extensions. Detailed histological analy-ses using light microscopy and marker gene stainingrevealed that #61 is required for the early choroid fis-sure formation. On the other hand, #61 may not bedirectly involved in the later eye development in-cluding retinal differentiation and optic stalk/nerveformations. In the #61-MO injected embryos, abnor-mal vessel structures in choroid fissure and trunkwere observed and hemorrhages were frequentlyseen at some parts of the eye and brain, indicating acritical role of this gene in the formation of the bloodvessel structures. Nevertheless, an early vascularmarker was found to be expressed properly and theestablishment of vascular network was also accom-plished normally in the #61-MO embryos. Takentogether, these results suggest that the novel gene#61 may play a key role in the maintenance of vascu-lar integrity during cardiovascular development inzebrafish embryos.

6. Preclinical studies using non human primateof Common Marmoset toward the therapeuticintroduction of human embryonic stem cellsinto clinical field

Erika Sasaki, Chieko Nakagawa, Sumiko Wa-tanabe, Kenzaburo Tani3: 3Department of ClinicalGenetics, Hematoloby/Oncology Research Hospi-tal, Medical Institute of Bioregulation, KyushuUniversity

Although the establishments of human embryonicstem (ES) cell lines have offered much hope bypromising to greatly extend the numbers and rage ofpatients who could benefit from transplant, in vivoexperiments are not adequate for human ES cellsfrom the ethical standpoint of view. We focused onin vivo studies using non-human primates as a prom-ising alternative system for preclinical studies ofhuman ES cells. We are using common marmosetswhich has several advantages as a model system. Wefirst established common marmoset (CM) ES cell cul-ture system and investigated exogenouse genetransfer methods suitable for CMES cells. The CMEScells are obtained from WiCell, USA. Since primateES cells including CMES cells exhibit spontaneousdifferentiation during the culture, we examined suit-able fetal bovine serum (FBS) and culture conditions.We found that Knockout Serum Replacement (KSR)and cynomolgus monkey ES cells splitting mediumshowed good results for CMES culturing. Then, wedetermined the suitable protocol to introduce exoge-nous DNA into CMES and found that green

fluorescent protein (GFP) gene expressing VSV-Gpseudotyped human immunodeficiency virus vectorunder EF-1α promoter and GFP gene could be trans-duced into CMES on mouse embryonic feeder cells.Undifferentiated CMES cell colonies stably express-ing GFP were obtained. Interestingly, when CMEScells were transduced with the HIV vector contain-ing other promoter such as CMV, PGK and CAG, allof the GFP expressing CMES cells were differentiat-ed. Our established GFP expressing CMES cell lineswould be an excellent tool to determine the best con-dition for CMES cells to differentiate intohematopoietic cells both in vitro and in vivo.

7. Nuclear and chromatin structure in embryonicstem (ES) cells

Noriko Sato and Ken-ichi Arai

In order to elucidate the mechanism how mouseembryonic stem (ES) cells maintain pluripotency andundergo symmetrical self-renewal in the presence ofLIF (leukemia inhibitory factor), we have investigat-ed the nuclear and chromatin structure unique toundifferentiated ES cell. Chromosomes as whole ge-netic elements, are not uniformly packed into the cellnucleus. The spatial arrangement and the higher-or-der structure such as the intra-nuclear location aswell as the internal organization of chromosome ter-ritories, should govern the chromatin function.Moreover, chromosomes are not constant structure,but they change their modes and dynamics in thecourse of development.

We have asked whether the nuclear positioning,the replication timing and the compaction rate ofspecific gene loci, whose developmental expressionsare tightly regulated, are influenced by LIF with-drawal (4 days). We have performed FISH withprobes to the Brachyury (mesodermal marker gene)and the Oct3/4 (undifferentiated ES marker gene)gene loci on chromosome 17 and analyzed them inBrdU positive nuclei. Regardless of culture condi-tions, both gene loci have exhibited neither anappreciable difference in their localization patternsnor heterochromatin association. In addition, bothalleles of two different gene loci have shown early- to- intermediate replication timing patterns. However,only in differentiated ES cells, they have shown anallelic difference in the distance between the two locion the same chromosome. It indicates that the large-scale compaction of the one allele may be inducedupon differentiation. We are currently studying onthe biological significance of these phenomena.

8. STAT6 and Allergic Diseases

Yumiko Kamogawa, Ken-ichi Arai

IL-4 and IL-13 are known to be important cytok-

119

ines that induce Th2 differentiation. The ligation ofcytokine IL-4 and 13 to their cognate receptors leadsto the activation of number of signaling pathwayswithin the cells. Activation of transcriptional factorSTAT6 is essential for the full response of cells tothose cytokines. The gene disruption study of STAT6showed that the abrogation of STAT6 leads to thefailure of Th2 differentiation and allergic responses.To elucidate the role of STAT6 in allergic diseasessuch as asthma, we have generated transgenic miceexpressing a conditionally active form of the protein(STAT6ER) by fusing STAT6 to the modified hor-mone-binding domain of estrogen receptor underthe CAG promoter. This protein was expressed inlungs, lymphocytes, heart, and brain and activatedby the addition of estrogen analog 4 hydroxy-tamox-ifen(4-HT) in vivo and in vitro. Activation ofSTAT6ER by 4-HT induced MHC class II upregula-tion in B cells and led to Th2 differentiation in T cellstogether with TCR stimulation in vitro.

Recently, it was reported that administration ofeither IL-4 or IL-13 cause asthma in the absence oflymphocytes in vivo. These reports suggested thatIL-4, 13 normally produced by Th2 cells directlystimulate lung epithelial cells and bronchial musclecells in lungs to generate asthmatic phenotype. Inaccordance with these results, activation ofSTAT6ER in vivo also caused airway hyper-respon-siveness and bronchial epithelial cell hyperplasia inthe absence of T and B cells in RAG backgroundtransgenic mice. This result indicated that the activa-tion of STAT6 in lungs is essential for initiation ofallergic asthma, therefore the molecules induced bySTAT6ER can be important candidates for asthma.To pursue the target molecules for causing asthma,we have currently investigated the molecules in-duced by 4-HT in transgenic mice lungs using bothRDA method and micro-array analysis.

9. Role of NFATx (NFAT4/NFATc3) in Expressionof Immunoregulatory Genes in Murine Periph-eral CD4+ T Cells

Jingtao Chen, Yumiko Kamogawa, ShoichiroMiyatake

Ca2+-regulated NFAT family members are tran-scription factors crucial for the expression of variouscytokine genes and other immunoregulatory genes.Analyses of mice defective in one or two NFAT fam-ily members have revealed functions specific to eachNFAT gene. However, the redundant functions ofseveral family members limit the usefulness of genedisruption analysis. For example, CD4+ T cells isolat-ed from NFATx-disrupted mice do not show anymodulation in cytokine gene expression, perhaps be-cause other family members compensate for itsabsence. In order to analyze the role of NFATx in theregulation of immunoregulatory genes in T cells, we

made a gain-of-function mutant by creating trans-genic mice expressing a constitutively nuclear formof NFATx in T cell lineages. The exogenously ex-pressed mutant localizes mainly in nucleus in theabsence of a Ca2+ signal. The transcriptional activityof this mutant measured by the reporter plasmid car-rying NFAT/AP-1 composite sites in the presence ofPMA is comparable to or higher than that of the wildtype NFATx stimulated by PMA and Ca ionophore.Thus, in addition to the effects of over-expression,the activity of this mutant can be distinguished fromthat of the endogenous wild type NFATx by compar-ing their responses under two different conditions;stimulation solely by PMA and stimulation by PMAand Ca ionophore. In naïve CD4+ T cells NFATx up-regulated the expression of several cytokine genes(IL-2) and activation markers such as CD25 (the IL-2receptor α chain), CD69 and CD62 ligand (CD62L)and suppressed the expression of CD154. In Th1cells, NFATx enhanced the expression of the Th1 cy-tokine genes, IFNγ and TNFα. In contrast, NFATxsuppressed Th2 cytokine genes such as IL-4 and IL-5in Th2 cells. It has been reported that both NFAT1and NFATx are required to maintain the homeosta-sis of the immune system. Our results suggest thatNFATx exerts this function by inhibiting the expres-sion of some critical immunoregulatory genes.

10.Analyses of Cdc7 conditional knockout EScells and mice

Jung Min Kim, Naofumi Takemoto4, Hiroko Fujii-Yamamoto, Ken-ichi Arai, and Hisao Masai5:4Department of Immunology, Tokyo MetropolitanInstitute of Medical Science, Tokyo 113-8613, Ja-pan, 5Department of Cell Biology, TokyoMetropolitan Institute of Medical Science, Tokyo113-8613, Japan

Cdc7 kinase is essential for initiation and progres-sion of DNA replication. Cdc7-/- mouse ES cells arenon-viable but their growth can be rescued by anextopically expressed transgene (Cdc7-/-tg). Condi-tional inactivation of Cdc7 in ES cells results inimmediate replication fork arrest, followed by p53-dependent cell death. S phase-arrested nucleiisolated from Cdc7-/- ES cells cannot be replicated invitro unless supplemented with cytosol from thewild-type cells.

Despite the normal growth capability of the Cdc7-

/-tg ES cells, the mice with the identical geneticbackground exhibit growth retardation. Further-more, Cdc7-/-tg mice display severe testicularhypoplasia and disrupted spermatogenesis at or be-fore the meiotic prophase I. The impairment inspermatogenesis correlates with the extremely lowlevel of Cdc7 protein in testis, and is rescued by in-troducing an additional allele of transgene, whichresults in increase of Cdc7 expression. The increased

120

Nakamura, T., Ouchida, R., Kodama, T., Kawashima,T., Makino, Y., Yoshikawa, N., Watanabe, S.,Morimoto, C., Kitamura, T., Tanaka, H. (2002)Cytokine receptor common beta subunit-medi-ated STAT5 activation confers NF-kB activationthrough IB-independent mechanism in murineproB cell line Ba/F3 cells., J. Biol. Chem. 277, 6254-6265

Zeng, R., Aoki, Y., Yoshida, M., Arai, K., Watanabe,S. (2002) Stat5b shuttles between cytoplasm andnucleus in cytokine-dependent and –independentmanner. J. Immunol. 168, 4567-4575

Watanabe, S., Murakami, T., Nakamura, T.,Morimoto, C., Arai, K. (2003) Human GM-CSF in-duces HIV-1 LTR by multiple signalling path-ways. Biochimie 84, in press

Mitsui, T., Watanabe, S., Taniguchi, Y., Hanada, S.,Ebihara, Y., Sato, T., Heike, T., Mitsuyama, M.,Nakahata, T., Tsuji, K. (2003) Impaired neutrophilmaturation in truncated murine G-CSF receptor-transgenic mice. Blood, in press

Akagawa, E., Muto, A., Arai, K., Watanabe, S. (2003)Analysis of the 5’ promoters for human IL-3 andGM-CSF receptor a genes. Biochem. Biophys. Res.Comm., in press

Kurita, R., Sagara, H., Aoki, Y., Link, B. A., Arai, K.,Watanabe, S. (2003) Suppression of lens growth byaA-crystallin promoter-driven expression of diph-theria toxin results in disruption of retinal cell or-ganization in zebrafish. Dev. Biol. In press

Iwasaki-Arai, J. , Iwasaki, H., Miyamoto, T.,Watanabe, S., Akashi, K. (2003) Enforced GM-CSFsignals do not support lymphopoiesis, but instructlymphoid to myelomonocytic lineage conversion.J. Exp. Med. in press

Yuyama N, Davies DE, Akaiwa M, Matsui K,Hamasaki Y, Suminami Y, Yoshida NL, Maeda M,

Pandit A, Lordan JL, Kamogawa Y, Arima K,Nagumo F, Sugimachi M, Berger A, Richards I,Roberds SL, Yamashita T, Kishi F, Kato H, Arai KI,Ohshima K, Tadano J, Hamasaki N, Miyatake S,Sugita Y, Holgate ST, Izuhara K. Analysis of noveldisease-related genes in bronchial asthma.Cytokine. 2002 Nov;19(6):287-96.

Chen, J., Amasaki, Y., Kamogawa, Y., Nagoya, M.,Arai, N., Arai, K. and Miyatake, S.: NFATx(NFAT4/NFATc3) in expression of immunoregulatory genesin murine peripheral CD4+ T Cells, J. Immunol. inpress

Kim, J. M., Nakao, K., Nakamura, K., Saito, I.,Katsuki, M., Arai, K., Masai, H.: Inactivation ofCdc7 kinase in mouse embryonic stem cells resultsin S phase arrest and p53-dependent cell death.EMBO J. 21, 2168-2179, 2002

Yamada, M., Sato, N., Taniyama, C., Ohtani, K., Arai,K. and Masai, H.: A 63 base-pair DNA segmentcontaining a Sp1 site but not a canonical E2F sitecan confer growth-dependent and E2F-mediatedtranscriptional stimulation of the human ASKgene encoding the regulatory subunit for humanCdc7-related kinase. J. Biol. Chem. 277, 27668-27681, 2002

Tanaka, T., Mizukoshi, T., Taniyama, C., Kohda, D.,Arai, K. and Masai H.: DNA binding of PriA pro-tein requires cooperation of the N-terminal D-loop/arrested-fork binding and C-terminalhelicase domains. J. Biol. Chem. 277, 38062-38071,2002

You, Z., Ishimi, Y., Masai, H., Hanaoka, F.: Roles ofMcm7 and Mcm4 subunits in DNA helicase activ-ity of mouse Mcm4/6/7 complex. J. Biol. Chem.277, 42471-42479, 2002

Tanaka, T., Taniyama, C., Arai, K., Masai, H.: AT-Pase/helicase motif mutants of Escherichia coli

Publications

level of Cdc7 also recovers the growth of Cdc7-/-tgmice, indicating that the developmental abnormali-ties observed in Cdc7-/-tg mice are due to insufficientlevel of Cdc7 protein. Our results indicate the re-quirement of a critical level of Cdc7 kinase fornormal mouse development and reveal its essentialroles in meiotic processes in mammals.

11.Molecular basis for recognition of arrestedDNA replication forks: a critical role of the 3'-terminus of nascent DNA chains

Taku Tanaka, Toshimi Mizukoshi6, Ken-ichi Arai,Daisuke Kohda4, and Hisao Masai4: 6BiomolecularEngineering Research Institute, Suita, Osaka 565-0874, Japan

Arrest of replication forks by various internal andexternal threats evokes a myriad of cellular reactions,collectively known as DNA replication checkpointresponses. In bacteria, PriA is essential for restora-tion of stalled replication forks and recombinationalrepair of double-stranded DNA breaks and is a can-didate sensor protein that may recognize arrestedforks. We discovered that PriA protein specificallyrecognizes 3'-termini of arrested nascent DNAchains at model stalled replication forks in vitro. Mu-tations in the putative “3'-terminus binding pocket”present in the N-terminal segment of PriA result infailure to bind to stalled replication fork structuresand loss of its biological functions. The results sug-gest a general mechanism by which stalledreplication forks are recognized by a sensor proteinfor checkpoint responses.

121

PriA protein essential for recombination-depen-dent DNA replication. Genes to Cells, in press,2003

Masai, H., Arai, K.: Cdc7 kinase complex: A key

regulator for initiation of DNA replication (Re-view with a cover photo). J. Cell. Physiol., 190, 287-296, 2002

department of basic medical sciences division of molecular ......imai, k., and saito, h. tgf-β...

Documents