KURENAI : Kyoto University Research Information Repository

KURENAI : Kyoto University Research Information Repository
Annual Report of the Institute for Virus Research, Kyoto
University, Volume.49, 2006( 遺伝子動態調節研究部門 遺伝
子情報解析研究分野 )
Issue Date
Annual Report of the Institute for Virus Research, Kyoto
University (2007), 49
Kyoto University
1) Interactions among proteins involved in translesion DNA synthesis: E. OHASHI,
Cellular DNAs in living organisms are constantly exposed to various genotoxic agents
of endogenous and exogenous origins. To cope with different DNA damages generated
by various genotoxins, cells are endowed with multiple mechanisms for repair and
tolerance of DNA damages. One of DNA damage tolerance mechanisms is translesion
DNA synthesis (TLS) by specialized DNA polymerases (mainly Y-family polymerases),
which is carried out in co-operations with various partner proteins. First, all Y-family
polymerases (Polη, Polι, Polκ and REV1) were reported to interact with PCNA
(Proliferating Cell Nuclear Antigen), a protein that is known as the sliding clamp to
increase the processivity of many DNA polymerases. Second, all Y-family DNA
polymerase has one or two copies of ubiquitin-binding motif and indeed, some of the
proteins (e.g., Polι) were shown to bind to ubiquitin. Furthermore, Polη was reported to
interact with Rad18, an ubiquitin ligase (E3 enzyme) that forms a complex with Rad6
(ubiquitin conjugation enzyme, E2). In DNA-damaged cells, PCNA is ubiquitinated by
the Rad6-Rad18 complex, resulting in increasing affinity to Y-family polymerases.
Finally, Polη, Polι, and Polκ bind to the C-terminal region of REV1 (as described
below in more details). Thus, these TLS proteins have various binding partners and the
differences in the affinity to such partners may determine which enzyme is recruited to a
site of DNA damage.
In order to compare the binding affinities of each Y-family enzyme to the partner
protein based on yeast two-hybrid assay, we made quantitative measurement of
β-galactosidase activity. Among Y-family DNA polymerases, Polι was found to
interact most strongly with either PCNA or ubiquitin. REV1, Polη and Polι interacted
with Rad18 at similar strength. Polκ interacted most weakly with either one of PCNA,
ubiquitin and Rad18. As for binding with REV1, Polη interacted most strongly and
Polι did second most.
Thus, Y family DNA polymerases show different affinities to distinct binding
partners. However, it is possible that DNA damage signaling may change the biding
affinities, for example, by phosphorylation and regulates the accessibility of the TLS
2) The mechanism of the interactions between hREV1 and three Y-family DNA
polymerases: E. OHASHI, T. HANAFUSA, K. KAMEI, and H. OHMORI
As described above, Polη, Polι, and Polκ interact with PCNA and have a sequence
similar to the consensus sequence of so-called PIP (PCNA interacting protein) box,
which is frequently represented by Q-x-x-(I,L,M)-x-x-F-F (x is any amino acid residue).
We and other groups have shown that Polη, Polι, and Polκ interact with a C-terminal
region of REV1, another member of the Y-family DNA polymerases. The same
C-terminal region of REV1 was also shown to interact with REV7, a non-catalytic
subunit of Polζ, which is another TLS DNA polymerase belonging to the B-family.
This has suggested us that REV1 plays a crucial role in cellular TLS events. Unlike for
the PCNA binding, Polη, Polι, and Polκ did not seem to have an apparent consensus
sequence for the REV1-interaction. However, our further experiments revealed that
short sequences less than 10 residues are sufficient for the REV1-interaction in all of the
three Pols. In addition, in all cases, two consecutive phenylalanine (FF) residues were
found to be critical for the REV1-interactions. In contrast to the PCNA-binding
sequence, no conserved amino acid residue is required for the N-terminal side of the FF
motif, but the presence of several residues at its C-terminal side is essential while the
sequence is not conserved. Using surface plasmon resonance method, we showed that
short synthetic peptides of the sequence derived from the REV1-binding site directly
bind the C-terminal region of REV1. Furthermore, such peptides did inhibit the binding
between purified REV1 and Polκ proteins.
To examine the biological relevance of the interaction between REV1 and Polκ, we
employed complementation assay using Polκ-deficient MEF (mouse embryonic
fibroblast) cells, which exhibit increased sensitivity to BPDE (benzo[a[pyrene
dihydroxydiol epoxide, the ultimate carcinogen of benzo[a[pyrene] and UV-irradiation.
We observed that the sensitivities to BPDE or UV were recovered by transient
expression of the wild-type of Polκ (in a fusion with GFP), but not by that of the
FF567-578AA mutant deficient for the REV1-binding.
Thus, the inteUaction with REV1 is essential for in vivo functioning of Polκ. The
biological meaning of the REV-interactions of the three Pols needs to be elucidated by
further investigations.
3) Interaction of hREV1 with hREV7 and Polκ: T. HANAFUSA, E. OHASHI, H.
HASHIMOTO1, and H. OHMORI (1 Yokohama City University)
Polζ is one of multiple DNA polymerases involved in translesion DNA synthesis
(TLS) and composed of two subunits; the REV3 catalytic subunit and the REV7
accessory subunit. The REV3 subunit show a significant similarity in the amino acid
sequence to the catalytic subunit of Polδ, the replicative DNA polymerase, so that Polζ
is classified into the B-family, while many of other TLS DNA polymerases are
classified inot the Y-family. The Saccharomyces cerevisiae (sc) REV7 subunit was
shown to increase the polymerase activity of the scREV3 catalytic subunit by 20- to
30-fold, but the exact function of the REV7 is not yet known. The human REV7 protein
(hereafter hREV7) has a significant similarity to the human MAD2 protein (hereafter
designated hMAD2A), a protein involved in spindle checkpoint.
For the reason,
hREV7 is also called hMAD2B (hMAD2L2 or hMAD2β). In fact, Xenopus laevis
REV7 was shown to interact with Cdh1-APC (Anaphase Promoting Complex),
paralleling the effect of MAD2A on Cdc20-APC. The hMAD2B has been reported to
interact with some other proteins such as MDC9 (a metalloprotease disintegrin) and
PRCC (papillary renal cell carcinoma). Moreover, a recent study reported that hREV7
interacts with Elk-1, which is a transcription regulator and targets on egr-1 or c-fos
following exposure of cells to DNA damaging agents, and acts to promote Elk-1
phosphorylation by the JNK MAP kinases. Thus, hREV7 may play a critical role in
linking TLS to cell cycle regulation or gene expression changes.
Previously, hREV7 was shown to interact with a C-terminal domain (CTD) of
hREV1 (approximately 100 amino-acid in the total 1251 residues). Recently, we and
other groups have reported that the same CTD of hREV1 interacts also with Polκ Polι
and Polη. To investigate how REV1 interacts with such many other DNA polymerases,
we examined the structural requirements of hREV1 for the binding to hREV7. Our
recent experiments revealed that short sequences less than 10 residues of the three Pols
are sufficient for the REV1-interaction, all containing the FF motif. However, hREV7
does not contain FF in the entire 211 amino acid sequence and furthermore, a previous
study showed that an almost entire region of hREV7 is required for its interaction with
Therefore, it seems likely that hREV1 CTD interacts with hREV7 and the
three Pols in different manners. Our results also suggested that a tertiary structure of
the hREV1 CTD is required for the interaction with hREV7. However, in a sharp
contrast to it, we have very recently found that a sequence of REV3 as short as 14
amino acids is sufficient for the interaction with hREV7. Thus, hREV7 seems to have
multiple phases to interact with different proteins. We are now trying to purify a
complex of hREV7 and hREV1 CTD for structural analysis.
4) Possible Mechanisms for the Polκ 3’-mRNA processing: Y. S. CHAN, H.
Polκ is one of the Y-family DNA polymerases that are involved in translesion DNA
synthesis (TLS). Polκ is able to accurately and efficiently bypass some DNA lesions
such as dG-N -BPDE, a bulky adduct generated by benzo[a]pyrene that is a potent
environmental carcinogen.
Both the human and mouse genes coding for Polκ
POLK for the human
gene and Polk for the mouse gene) are ubiquitously expressed with the highest
expression in testis.
Besides Polk, many genes involved in DNA repair and
recombination functions are also observed to show similar expression patterns.
However, one unique feature of the Polk genes is that the major transcripts found in the
testes are shorter than those found in other organs; mRNAs abundantly expressed in the
testes are about 2.8 kb in length whereas those ubiquitously expressed in many other
organs (including testis at lower level) are 4.2 kb. The major and shorter transcripts
found in testes have a polyA tail immediately downstream of the translation stop codon,
whereas the longer transcripts have long 3’-UTR containing multiple copies of AUUUA
sequence (an essential and minimal unit of AREs). AREs are known to render mRNAs
unstable when present in 3’-UTR. Most of the tissues may express these labile mRNAs
to keep the amount of Polk mRNAs at low levels to avoid unfavorable mutations caused
by the accumulation of this error-prone enzyme.
Some mouse genes are also known to generate shorter transcripts in testes than in
other organs. One of such genes is Cdcc89, whose expression is dramatically
up-regulated during spermiogenesis. The Cdcc89 transcripts found in the testis are
about 1.5 kb in length while the transcripts found in embryonic brain are 2.3 kb.
Cdcc89 is an intron-less gene, and the two kinds of transcript have the same coding
region, differing in the length of 3’-UTR. It seems plausible that a same mechanism
operates to generate shorter transcripts of the Polk and Ccdc89 genes, making shorter
CstF-64 is a protein essential for polyadenylation of mRNAs, involved at the
cleavage process of nascent transcripts in the downstream of the canonical polyA signal
sequence AAUAAA. Recently, it has been found that a new form of CstF-64 (named
τCstF-64) is most highly expressed in male germ cells, which may cause the differences
polyadenylation site selections between somatic cells and male germ cells. We are
planning to test whether τCstF-64 may be involved in making shorter Polk and Ccdc89
transcripts in testes.
M. Kawanishi, K. Matsukawa, E. Ohashi, T. Takamura, Y. Totsuka, M. Watanabe, T.
Sugimura, K. Wakabayashi, F. Hanaoka, H. Ohmori, T. Yagi.: Translesion
DNA synthesis across mono ADP-ribosylated dG by Y-family DNA
polymerases, New Developments in Mutation Research, in press
X. Bi, L.R. Barkley, D.M. Slater, S. Tateishi, M. Yamaizumi, H. Ohmori, C. Vaziri.:
Rad18 regulates DNA polymerase κ and is required for recovery from S-phase
checkpoint-mediated arrest. Mol. Cell. Biol. 26, 3527-3540, 2006
㞕㸯7*; ࡛ࡡ┞பష⏕ࢅᾐ኶ࡊࡒ SROȞࡡࣃࢹ⣵⬂හ࡚ࡡよᯊࠉ᪝ᮇฦ
Ꮔ⏍∸Ꮥఌ ࣆ࢚࣭࣑ࣚࠔฦᏄ⏍∸Ꮥࡡᮅᮮࠕࠉྞཿᒁࠉ㸦㸧᭮㸫ࠤ
Department of Genetic and Molecular Biology
Laboratory of Gene Information Analysis
ᚋᚃࡵᏕ⾙ᣲ⮾ఌ≁ื◂✪ဤ PD ࡐࡊ࡙ຐᡥ୩ࡲᙽ㐕ࡡ◂✪ဤ࡛ࡊ࡙ᅹ⡘ࡊ࡙
Ꮥ⏍ Chan Yuet Sim ࠿⌦Ꮥ◂✪⛁⏍∸⛁Ꮥᑍᨯࡡಞኃㄚ⛤ M1 ࡛ࡊ࡙ᠻࠍࡡࢡࣜ
ᠻࠍ࠿ࡆࡆᩐᖳ୯ᚨㄚ㢗࡛ࡊ࡙᡽ࡖ࡙ࡀࡒ᥾ഭࣁ࢕ࣂࢪ㓕⣪ Polκࢅࢤ࣭ࢺ
ࡌࡾ㐿ఎᏄ㸝࣏ࢗࢪࡡ㐿ఎᏄࡢ Polkࠉࣃࢹࡡ㐿ఎᏄࡢ POLK ࡛⾪オࡈࡿࡾ㸞ࡡ
DNA ྙᠺࠉ⤄ࡲᥦ࠻ࠉಞᚗ࡝࡜࡞㛭ࢂࡾ㐿ఎᏄࡡኣࡂࡢ⢥ᕛ࡚ࡡⓆ⌟࠿㧏࠷ࡆ
࡛࠿▩ࡼࡿ࡙࠷ࡾ࠿ࠉࣃࢹ POLK ࡷ࣏ࢗࢪ Polk ࡡ୦㐿ఎᏄࡡⓆ⌟࡞ࡗ࠷࡙ࡡࣗ
ࢼ࣭ࢠ࡝≁㛏ࡢࠉ⢥ᕛ࡚ࡡ୹さ࡝ mRNA ࡢ⣑ 2.8 kb ࡡ㛏ࡈ࡚࠵ࡾࡡ࡞ᑊࡊ࡙௙
ࡡ⮒ჹ࡚Ⓠ⌟ࡈࡿ࡙࠷ࡾ mRNA ࡡ㛏ࡈࡢ⣑ 4.8 kb ࡛ࠉ⢥ᕛ࡚ࡡ mRNA ࡡ㛏ࡈ࠿
௙ࡡ⮒ჹ࡞Ẓ࡬࡙㢟ⴥ࡞▯࠷ࡆ࡛࡚࠵ࡾࠊᠻࠍ࡛ྜྷ᫤᭿࡞ POLK, Polk 㐿ఎᏄࢅ
ྜྷᏽࡊࡒ⡷ᅗࡡ E. Friedberg ࡡࢡ࣭ࣜࣈࡢ 4.8 kb ࡛㛏࠷㌹෕∸ࡡ 3’-UTR ࡞ࡢ
mRNA ࢅ୘Ꮽᏽ࡞ࡌࡾ ARE (AU-rich element)࠿ኣࢤࣅ࣭Ꮛᅹࡌࡾࡆ࡛ࢅሒ࿈ࡊ
ࡒࠊᠻࠍࡡࢡ࣭ࣜࣈࡢࣃࢹ POLK ࡷ࣏ࢗࢪ Polk ࡡ୦㐿ఎᏄࡡ㌹෕㛜ጙⅤࢅḿ☔
࡞ሲᇱ࡚ࣝ࣊ࣜỬᏽࡊ࡙ࠉ㸦㸞࣏ࢗࢪ Polk 㐿ఎᏄࡡ㌹෕ࡢ஦ࡗࡡ␏࡝ࡾన⨠
ࠉཀྵࡦ P1b)࠾ࡼ㛜ጙࡈࡿࠉ୕Ὦ࡞న⨠ࡌࡾ P1a ࠾ࡼࡡ㌹෕㛜ጙࡢ⢥ᕛࢅ
ྱࡳኣࡂࡡ⮒ჹ࡚㉫ࡆࡾ࠿ࠉࡐࡿ࠾ࡼ⣑200ࢽࢠ࢛ࣝࢲࢺୖὮ࡞Ꮛᅹࡌࡾ 1 ࡡ
㌹෕㛜ጙࡢ࡮࡯⢥ᕛ≁␏Ⓩ࡚࠵ࡾࡆ࡛ࠉ㸧㸞࣏ࣃࢹ POLK 㐿ఎᏄࡡ㌹෕㛜ጙ௛
㎾ࡡ DNA ሲᇱ㒼าࡢ୕オࡡ࣏ࢗࢪࡡ㒼า࡛㢦జࡊ࡙࠷ࡾ࠿ࠉࡐࡡ㌹෕㛜ጙࡢ࣏
ࢗࢪࡡ P1a ࡡ 6 ࢽࢠ࢛ࣝࢲࢺࡡୌ⟘ᡜ࠾ࡼࡡࡲ㉫ࡆࡾࡆ࡛ࢅ♟ࡊࡒࠊୌ᪁ࠉ࣏
ࢗࢪ Polkࠉཀྵࡦࣃࢹ POLK ࡡ⢥ᕛ࡞࠽ࡄࡾ mRNA ࡢ⩳ゼ⤂ጙࢤࢺࣤࡡࡌࡃୖὮ
࡚ polyA 㒼า࡞⦽࠿ࡾࡡ࡞ᑊࡊ࡙ࠉ௙ࡡ⮒ჹ࠾ࡼࡡ mRNA ࡢ㛏࠷ 3’-UTR ࢅᣚ
ࡖ࡙࠷ࡾࠊࡌ࡝ࢂࡔࠉ⢥ᕛ㸡ཀྵࡦ௙ࡡ⮒ჹ࡚షࡼࡿࡾ mRNA ࡢ coding region ࡢ
ධࡂྜྷࡋ࡚࠵ࡾ࠿ࠉ3’-UTR ࡞࠽࠷࡙ኬࡀࡂ␏࡝ࡾࠊ
ఎᏄ⩄ (TISP, transcript induced in spermiogenesis)ࢅྜྷᏽࡊ࡙ࠉࡐࡡ࠹ࡔࡡ࠵ࡾࢡ
࣭ࣜࣈ(Type 3ࠉTISP201~232)࡚ࡢ⢥Ꮔᙟᠺࡡᠺ⇅໩࡞ఔࡖ࡙ mRNA ࡡࢦ࢕ࢫ࠿
ᑚࡈࡂ࡝ࡾࡆ࡛ࢅ᪺ࡼ࠾࡞ࡊࡒࠊᐁ㝷ࠉTISP201~232 ࡡࡐࡿࡑࡿࡡධ㛏ࡡ cDNA
࡛ EST (expression sequence tag) library Ⓡ㘋ࡈࡿ࡙࠷ࡾ⢥ᕛ⏜ᮮࡡ cDNA ࡡࢦ࢕ࢫࠉ
ཱིࡽฦࡄ 3’-UTR ࡡࢦ࢕ࢫ࡞Ἰ┘ࡊ࡙Ẓ㍉ࢅࡌࡾ࡛ࠉTISP204 ࡚ࡢ⢥ᕛ࡚Ⓠ⌟ࡈ
ࡿࡾ mRNA (ḿᘟ㐿ఎᏄྞࡢ Ccdc89, coiled coil domain containing 89) ࡢ௙ࡡ⮒ჹ
㸝ࡆࡡሔྙࡢ⫶ඡࡡ⬳㸞ࡡ mRNA ࡻࡽࡵ▯࠷ 3’-UTR ࢅᣚࡗࡼࡊ࠷ࡆ࡛࠿ฦ࠾
ࡖࡒࠊࡐࡆ࡚⏛୯ࡼ࡛භྜྷࡊ࡙ࠉ⢥ᕛཀྵࡦ⫶ඡ࡚Ⓠ⌟ࡈࡿ࡙࠷ࡾ Ccdc89 ࡡ cDNA
ࢅ༟㞫ࡊ࡙ሲᇱ㒼าࢅよᯊࡊࡒ࡛ࡆࢀࠉ⫶ඡ⏜ᮮࡡ Ccdc89-cDNA ࡢ㸦 kb ࡮࡜
ࡡ 3’-UTR ࢅᣚࡔࠉࡐࡡ 3’-end ࡡ polyA 㒼าࡡ 20 ࢽࢠ࢛ࣝࢲࢺ୕Ὦ࡞ AGTAAA
࡛࠷࠹㒼า࠿Ꮛᅹࡊࡒ࠿ࠉ⢥ᕛ⏜ᮮࡡ Ccdc89-cDNA ࡡ 3’-UTR ࡢ⣑ 120 ࢽࢠࣝ
࢛ࢲࢺ࡛▯ࡂࠉࡐࡡ polyA 㒼าࡡ 20 ࢽࢠ࢛ࣝࢲࢺ୕Ὦ࡞ࡢ AGTACA ࡛࠷࠹㒼
า࠿Ꮛᅹࡊࡒࠊࡆࡡ Ccdc89 㐿ఎᏄ࡞ࡢ intro ࡢᏋᅹࡎࡍࠉ3’-UTR ࡡ㐢࠷ࡢ
alternative splicing ࡞ࡻࡾࡵࡡ࡚ࡢ࡝ࡂࠉpolyA site ࡡ㐽ᢝࡡ㐢࠷࡞ࡻࡾ࡛⩻࠻ࡼ
┷ᰶ⏍∸࡞࠽ࡄࡾ mRNA ࡡ 3’-end ࡢୌ᪞㌹෕ࡈࡿࡒᚃ࡞ AAUAAA ࡛࠷࠹
㒼าࡡ⣑㸧㸥ࢽࢠ࢛ࣝࢲࢺୖὮ࡚ว᩷ࡈࡿ࡙ࠉpolyA ࠿௛ຊࡈࡿࡾࡆ࡛࠿▩ࡼࡿ
࡙࠷ࡾࠊࡐࡡࡻ࠹࡝ว᩷ཬᚺ࡞㛭ࢂࡾᅄᏄࡡୌࡗ࡚࠵ࡾ CstF-64 ࡢ X ᯹Ⰵమ୕
࡞Ꮛᅹࡌࡾࡡ࡚ࠉ⢥Ꮔᙟᠺࡡ㏭୯ࡡṹ㝭࡚ࡐࡡⓆ⌟࠿ೳḾࡊࠉࡐࡡ௥ࢂࡽ࡞ 10
␊᯹Ⰵమ୕࡞ᏋᅹࡌࡾτCstF-64 ࠿⢥ᕛ≁␏Ⓩ࡞Ⓠ⌟ࡈࡿࡾࠊࡆࡡτCstF-64 ࠿഼ࡂ
࡛ AAUAAA ௧አࡡ㒼าࢅ polyA ࢨࢡࢻࣜ㒼า࡛ࡊฺ࡙⏕ࡌࡾྊ⬗ᛮ࠿⩻࠻ࡼࡿ
ࡾࠊ௑ᚃࡢ PolkࠉCcdc89 ࡛࠷࠹஦ࡗࡡ㐿ఎᏄࡡ⢥ᕛ࡞࠽ࡄࡾ mRNA ࡡ 3’-end
processing ࡞ࡗ࠷࡙ࡡ◂✪ࢅ᥾ഭࣁ࢕ࣂࢪࡡ࣒࢜ࢼࢫ࣑ࡡ◂✪࡛୩⾔ࡊ࡙㐅ࡴ࡙