- Email: [email protected]
DNA polymerase ε: in search of a function
TIBS 17 - FEBRUARY1992 Communication is another area of difficulty. Problems include the wide geographical spread of member countries, and, in some countries, poor communication facilities or the lack of funds for sending a representative to Council Meetings. Council Delegates are the lifelines that link the central organization of the FAOB to Constituent Members and unless that lifeline is effective, then that society and its members will become isolated from the FAOB. In an effort to improve communications, the FAOB issues a biennial newsletter for each member society to photocopy and distribute to its members. The President also sends a monthly letter to Council delegates,
Other forms of publication have been recently discussed. The launch of an FAOB research journal was originally debated over ten years ago but discussion has now broadened to include other kinds of publication, such as a review-type publication or a members' magazine. The primary objectives of the FAOB in launching a new publication are still a topic of debate: should it increase international prestige, generate income, or create a vehicle for enhancing communication and promoting a sense of identity in members? It is hoped that some consensus on this issue will soon be reached. The decision is an important one. The launch of any publication will reflect how
Constituent Members view the future role of the FAOB in the region over the next few years. Having witnessed much of the progress that has been achieved over the years, I still feel that the essential spirit of the FAOB has remained the same, namely that it is a Federation that exists for and exists through the cooperation of Constituent Members. If all those involved in the FAOB can retain this spirit of cooperation, then I am sure that the FAOB can find solutions to its immediate problems and have an even greater impact on the development of biochemistry in the AsianPacific region in the third decade of its existence.
TALKINGPOINT ALL PROKARYOTIC and eukaryotic cells contain several DNA polymerases. Besides replication of the chromosomal DNA, the cell requires DNA polymerases for other DNA synthesis events such as DNA repair, nuclear DNA recombination and replication of extrachromosomal (i.e. plasmid or mitochondrial) DNA. In bacteria, these tasks are shared by the
three different DNA polymerases I, II and III. In some instances plasmids and bacteriophages (e.g. T4 or T7) encode their own DNA polymerases ]. At least five different DNA polymerases have been identified and purified from eukaryotic cells 2. The subunit composition of the yeast enzymes is nearly identical to the one in higher eukaryotes which has lead to a new nomenclature for eukaryotic DNA polymerases 3. According to this new classification, all eukaryotic polymerases are denoted by greek letters (Table I).DNA polymerase (~ (reviewed in Ref. 4) is important in nuclear DNA replication. Recent data strongly suggest that it has a role in the initiation of the lagging strand of the replication fork 5. DNA polymerase ~ (reviewed in Ref. 6) is the major repair
DNA polymerase e: in search of a function .
.
.
.
.
.
.
The current model of eukaryotic DNA replication involves the two DNA polymerases 8 and (x as the leading and lagging strand enzymes, respectively. A DNA polymerase first discovered in yeast has now been found in all eukaryotic cells and is termed DNA polymerase E. In yeast, the gene for DNA polymerase E has recently been found to be essential for viability, raising new questions about its functions.
enzyme in the nucleus and appears to be involved in DNA recombination 7. DNA polymerase y (reviewed in Ref. 6) carries out DNA replication of the mitochondrial genome. DNA polymerase 8 (reviewed in Refs 8 and 9), first described in 1976, was originally distinguished from DNA polymerase ~ by t~ v its 3 5 proofreading exonuclease activity ]°. Subsequently, all eukaryotic 3'-~5' exonuclease-containing DNA polymerases were called DNA polymerase 8. DNA polymerase 5 is involved in nuclear DNA replication and is the Ulrich Hiil~©her and Pia Th6mmes are at the prime candidate for replication of the Department of Pharmacologyand Biochemistry, Universityof Z0rich-lrchel, leading strand of the replication fork. Winterthurerstrasse 190, CH-8057 Z0rich, On model homopolymer template/ Switzerland. primers [e.g. poly(dA)/oligo(dT)] con© 1992, ElsevierScience Publishers, (UK) 0376-5067/92/$05.00
taining long single-stranded regions, this enzyme requires an auxiliary protein, called proliferating cell nuclear antigen (PCNA), for processive DNA synthesis8. In higher eukaryotic cells, a second 3'45' exonuclease-containing DNA po]ymerase was isolated from calf thymus n and HeLa ceils ]2 which was independent of PCNA for processive DNA synthesis. A similar enzyme had been discovered more than 20 years ago in yeasP 3. Originally known as DNA polymerase II (Table ID, this enzyme is now called DNA polymerasec (reviewed in Ref. 14). Table II gives a brief history of DNA polymerases (~, 8 and ~. What is the function of DNA polymerase E? An involvement in the
55
TIBS 17 - FEBRUARY 1 9 9 2 and Table I). DNA pol~nerase ~ is com-
Table I. The five eukaryotic DNA polymerases a
DNA Yeast Biological polymerase gene function
(z
Catalytic subunit (kDa)
Other Function polypeptides (kOa)
posed of four polypeptides of approximately 160-185, 70, 60 and 50 kDa 4, and DNA polymerase 8 contains two polypeptides of 125 and 50 kDa 9. The sub-
160-185
70 50/60
40
-
unit structure of DNA polymerase E is still uncertain. Studies from yeast and calf thymus suggest that two forms exist (reviewed in Ref. 9). One form
Replicationof mitochondrial DNA POL3 Replicationof leadingstrand
125
50
Unknown
125
48
Unknown
POL2 Repairof nuclear DNA,
210-230 or 125-140
29/30/ 34/80 or 4O
All unknown
POLl
Replicationof laggingstrand, initiation of leading strand, repair ofnuclearDNA
J3
-
T
MIP1
Repairof nuclearDNA, recombination
replication of nuclear DNA
Anchor?. Primase heterodimer
(called DNA polymerase e* and, earlier,
asee Ref. 3 for a revisednomenclatureof eukaryotic DNA polymerases.Furtherreferencesare cited with-
in the text.
II in yeast) has a catalytic subunit of 125--140 kDa 23 and possibly one smaller subunit of 40 kDa TM, while the other form (called DNA polymerase ~ or earlier If* in yeast) is more complex with a catalytic subunit of 215-230 kDa and several subunits of between 30 and 70 kDa 24. The role of these two forms is not yet clear but they both exist in crude
Year
Protein
Comment
Ref.
extracts from yeast 24 and calf thymus (T. Weiser and U. HObscher, unpublished data). The sequence of yeast DNA polymerase c predicts a catalytic subunit of >250 kDa 15. Genetic experiments with the DNA polymerase c gene, however, indicated that the C-terminal half of DNA polymerase ~ was not essential for growth ~5. The N-terminal part coding for a 130 kDa DNA polymerase e was enough to restore viability of the cells, although with a slower growth rate. The catalytic core of mammalian DNA polymerase e (258 kDa) might be composed of two segments that are linked via a proteasesensitive area 23, thus resulting in either an enzyme of >200 kI)a or of 140 kDa. The 140 kDa DNA polymerase c is able to replicate a primed M13 DNA efficientlyTM. However, the mode of replication by DNA polymerase e appears to be more complex. The yeast 2s,26 and the human 27 enzymes become inactive upon addition of physiological salt concentrations. Replication activity is only restored if DNA polymerase ~ is supplemented with auxiliary proteins such as replication factor A (RF-A), PCNA and replication factor C (RF-C)2s,26. Replication of primed M13 in vitro is achieved by yeast DNA polymerase c,
1957
pol a a
Discoveryin calfthymus
28
together with those three auxiliary pro-
1970
pol ~b
Discovery in yeast
1976 1985 1986 1987 1988 1989 199o 1990
pol 6 yeast pol ac PCNA" PCNA pol E yeast pol ~
Discoveryin calf thymus Essentialgene function Auxiliaryproteinfor pol 5 Essentialfor SV40 replication in vitro PCNA-independent form Es.~entialgene function New nomenclature Esser)tialgene function
13 10 29 16 17 11 30 3 15
teins, at a speed that reaches that 0bserved in vivo2S,2E For this the 140 kDa proteolysed subunit of DNA polymerase
repair of nuclear DNA is suggested by its ability to complement in repair assays. However, that its gene is essential for yeast viability suggests an important role in DNA replication. How does this fit into our current image of the replication process?
DNA repair system yielded a iS-like DNA polymerase which was independent of PCNA~9. Based on an agreement reached at the Cold Spring Harbor Symposium in 1989, the PCNA-independent DNA polymerase 8 is now called DNA polymerase ca. For some time it was not possible to isolate DNA polDNA polymetaseE ymerases 8 and c from the same tissue, The finding that PCNA is an auxiliary Recently this has been achieved in protein of DNA polymerase 8 ~6and also HeLa cells 2°, human placenta 2~ and calf an essential factor for SV40 replication thymus 22. Biochemical comparison in vitro n brought about a renaissance of clearly distinguished DNA polymerase ~ interest in DNA polymerase 6. There from DNA polymerases 8 and (~/primase. was initial controversy, however, when The catalytic subunit of DNA polanother DNA polymerase was dis- ymerase ~ has been cloned and secovered u,18 that was independent of quenced from the yeast Saccharomyces PCNA for processive synthesis on cerevisiae. The gene was found to be model homopolymer templates such as essential for yeast viability, and a funcpoly(dA)/oligo(dT). This so-called PCNA- tion for this third essential DNA polindependent DNA polymerase 8 also ymerase in replication was therefore contained an associated proofreading suggested 15. 3'45' exonuclease activity. At about the DNA polymerases (z, 8 and ~ have difsame time, fractionation of an in vitro ferent subunit compositions (Refs 20-22
Table 2. Chronological history of the discovery of DNA polymerases ~, ~ and E
all pols
yeast pol E
a pol, DNA potymerase,
b Yeast DNA potymeraseEwas called DNA polymeraseII until 1990. c Yeast DNA polymerasea was called DNA polymeraseI until 1990. d PCNA,proliferatingcell nuclearantigen.
56
g was sufficient (P. M. J. Burgers, pers. commun.). In the SV40 in vitro replication systern ~7, DNA polymerase ~ from HeLa cells could not substitute for DNA polymerase 6. This enzyme was composed of a catalytic subunit of 210 kDa. If, however, the yeast DNA polymerase (likely a proteolysed 140 kDa E* form)
TIBS 17 - FEBRUARY1992
was used, a partial replacement of DNA polymerase 8 was observed. Since the high molecular weight DNA polymerase was originally isolated as a repair activity~9and the proteolysed DNA polymerase e* form can efficiently replicate in vitro, we suggest that proteolysis might not be merely an artifact of isolation, but might enable different functional forms to exist in the cell.
Howmightthree DNApolymerasesact at the replication fork? A model originally proposed by Sinha et a l ) 1 predicted that replication of the
leading and lagging strand might be performed by two different polymerase activities which have to be coordinated in some way. According to this model (Fig. 1), one DNA polymerase molecule replicates the leading strand continuously and the other replicates the lagging strand discontinuously. The lagging strand bends back by 180° on itself to form a loop at the replication fork. This would allow the dimeric DNA polymerase to move along both template strands in the 'same' direction without violating the 5'~3' directionality rule. When the lagging strand DNA polymerase reaches the Okazaki fragment synthesized during the previous round, the freshly synthesized DNA is threaded through the enzyme, allowing the DNA polymerase to recycle to a newly exposed single-stranded region placed 3' to the previous priming site for initiation of the next Okazaki fragment. In this way, a dimeric DNA polymerase holoenzyme would guarantee an efficient and coordinated progression of the advancing replication fork. Data supporting this model accumulated in the T4 DNA replication system 3z, for the Escherichia coli DNA polymerase III holoenzyme33,34and for various forms of the DNA polymerase holoenzyme from calf thymus3E In E. coli this dual DNA polymerase funcffOn appears to be founded on the asymmetry of the DNA polymerase III holoenzyme3E Three subunits of the DNA polymerase III core polypeptides are identical in both the leading and the lagging DNA polymerase, but each complex is associated with a variety of different additional subunits. The composition of these subunits provides each half of this twin DNA polymerase III with properties suited for replication of either the leading and the lagging strand. DNA polymerase IIl was isolated as an asymmetric dimer with 22 polypeptides 37 of which one subunit
(the x-subunit) appears to s' s' have the capacity to dimerize the DNA polymerase IIl core ~. The ~!'~ t structural basis of the .. ,~ asymmetric dimer DNA polymerase III may be due ......~ . . . _ to the occurrence of two . . ~ '.'~" s' related but differentially ~ . . processed accessory poly~ymerase~ "'-. "~.~ peptides in the same com1 ~ _ 'a°g~ng i:~ "'..... Y~_ plex (see, for example, '~ .:':~.;~ :~ , ' : " • Refs 38 and 39). ~:i~,,~6,~_,~J Work by Stillman et al. 5,4° indicated that sequential initiation of leading and lagging strand synthesis at the replication origin of SV40occurs by two different DNA polymerases, and 8. After the origin of Figure 1 The asymmetric DNA replication fork. Most organisms replication is unwound by require an asymmetrictype of DNAreplication fork for the combined action of T replicationof both DNAstrands. Dotted lines, parental antigen and the singlestrands; solid lines, newlyreplicated DNA;black bars, stranded DNA binding primers. protein (SSB, also called RF-A or RP-A), DNA polymerase a/primase initiates DNA syn- of RF-C to a reaction mixture containing thesis of the lagging strand (see Refs 41 DNA polymerase ~, PCNA and RF-A. The and 42 for a recent review on enzymes high degree of functional similarities and proteins acting at the replication between DNA polymerase 5 and ~ in fork). When a short stretch of DNA has their capacities to form a holoenzyme been formed, the two DNA polymerase structure with the same auxiliary proauxiliary proteins PCNA and RF-C come teins zS, and the fact that, in yeast, both into action by forming a tight complex enzymes are essential gene products 1s,3°, at the 3'OH end of the growing DNA suggest that both DNA polymerases chain in the presence of ATP. This bind- might function in chromosomal DNA ing prevents DNA polymerase a/pri- replication. mase from further binding to the 3'OH A prospective working model is outend and instead allows DNA polym- lined in Fig. 2. After unwinding at a erase ~i to bind and start synthesis of given origin of replication, the initiating the leading strand. The dissociated DNA polymerase (DNA polymerase DNA polymerase ~t/primase cycles to a/primase) synthesizes primers and the next priming site on the lagging short pieces of DNA for lagging strand strand to resume discontinuous repli- synthesis at both forks. The primase cation. Simultaneously, DNA pol- tightly associated with DNA polymerase ymerase ~ continues to synthesize the a makes it an indispensible enzyme for leading strand. The lagging strand syn- priming events at the lagging strand. thesized by DNA polymerase c~ serves Following this initiation of synthesis, to initiate the synthesis of the leading DNA polymerase a/primase is replaced strand, subsequently carried out by DNA by other DNA polymerases for the elonpolymerase ~iand its auxiliary proteinsE gation mode (for DNA polymerase How does DNA polymerase E, the switch, see Ref. 5). Figure 2 depicts how third essential DNA polymerase come three DNA polymerases might share the into play? DNA polymerase auxiliary replication tasks during the elongation proteins, such as RF-A, RF-C and PCNA, event: the leading strand is replicated are required by DNA polymerase ~5for by a processive DNA polymerase (DNA replication of natural DNAs,43. These polymerase ~5or DNA polymerase e). At three proteins also appear to interact the same time the initiating DNA polwith DNA polymerase E but probably in ymerase (DNA polymerase a/primase) a more complex way than with DNA continues to synthesize primers and polymerase ~i25,26. As mentioned above, short pieces of DNA at the lagging DNA replication by DNA polymerase e strand. Subsequently, the Okazaki fragbecomes resistant to salt upon addition ments are synthesized to completion by
%
57
TIBS 17 - F E B R U A R Y 1 9 9 2
Initiation
,"
... •.""
.....................................' , , . . . . .
%
". ''.. ~ % ~
~®
............ s' J I ~( ( ( [ 3'.
; - ................. 5' LI..I..J..~..~.~.'[3'
""".,° ". ....•. . .
..-"
.......... ..............................
Elongation
~
' ."
l ..................................... " r . , . . .... ...." •" "
! ~ "'"
.* •."
.." ../ •'
: ©~ "=" " C ,~
......
~~~~
"
io.~
_c'~l.......
[
~
~$ ~
:#j~
..'K~,~ "
"::'Y" "'"'"
~
I ~ .=
~
.......". ....
~:
".~:...........................
~~5,
"T':f'j'5
~
-a ~ . . ~ ~
~
~,..,"~g,~ ~" ~ ~
~ ~ . , . . ~ - - , , / ~ , ~ , ~
~';,'f
"
a" ~ ¢ ~ :~. ~~ :~ ~"
.~.....-
~ m /
-
191, 617-618 4 Lehman, I. R. and Kaguni, L. S. (1989) J. Biol. Chem. 264, 4591-4595 5 Tsurimoto, T., Melendy, T. and Stillman, B. (1990) Nature 346, 534-539 6 Fry, M. and Loeb, L. A. (1986) Animal Cell DNA Polymerases, CRC Press Inc. 7 Nowak, R., Woszczynski, M. and Siedlecki, J. A. (1990) Exp. Cell Res. 191, 51-56 8 Downey, K. M., Tan, C-K. and So, A. G. (1990) BioEssays 12, 231-236 9 Bambara, R. A. and Jessee, C. B. (1991) Biochim. Biophys. Acta 1088, 11-24 10 Bymes, J. J., Downey,K. M., Black, V. L. and So, A. G. (1976) Biochemistry 15, 2817-2815 11 Focher, F. et al. (1988) Nucl. Acids Res. 16, 6279-6295 12 Syv~oja,J. and Linn, S. (1989) J. Biol. Chem. 264, 2489-2497 13 Wintersberger, U. and Wintersberger, E. (1970) Eur. J. Biochem. 13, 11-19 14 Suv~oja, J. (1990) BioEssays 12, 533-536 15 Morrison, A. et al. (1990) Cell 62, 1143-1151 16Tan, C-K., Castillo, C., So, A. G. and Downey, K. M. (1986) J. Biol. Chem. 261, 12310-12316 17Prelich, etal. (1987)Nature326, 471-475 18 Focher, G. F. et al. (1989) Nucl. Acids Res. 17, 1805-1821 19 Nishida, C., Reinhard, P. and Linn, S. (1988) J. Biol. Chem. 263, 501-510 20 Syv~oja,J. et al. (1990) Prec. Natl Acad. Sci. USA 87, 6664-6668 21 Lee, M. Y. W. T., Jiang, Y., Thang, S-J. and Toomey, N. L. (1991) J. Biol. Chem. 266, 22 Weiser, T. et al. (1991) J. Biol. Chem. 266,
10420-10428 23 Kesti, T. and Syv,oja, J. (1991)J. Biol. Chem.
;.."
266,6336-6341 24 Hamatake, R. K. et al. (1990) J. Biol. Chem.
265, 4072-4083 25 Burgers, P. M. J. in Eukaryotic DNA Polymerases
Figure 2 Model for eukaryotic DNA replication, depicting how three DNA polymerases might participate in DNA replication. For explanation see text. Dotted lines, parental strands; solid lines, newly replicated DNA; black bars, primers.
a third DNA polymerase (DNA polymerase e or DNA polymerase ~). It is conceivable that various subsets of auxiliary protein compositions (PCNA, RF-A, RF-C and others, such as DNA hellcases) might tune the DNA polymerases and e to their respective functional roles. This model would explain how proofreading of the two strands might occur since both DNA polymerases 8 and c contain an intrinsic 3'~5' exonuclease9.
appear to form holoenzymes with the same auxiliary subunits and replicate DNA in v i t r o at physiological rates. The functional similarity of DNA polymerases and their auxiliary proteins from yeast to man strongly argues for the conservation of the eukaryotic rep]ication fork. Acknowledgements
We thank C. C. Kuenzle and M. Berchtold for critical reading of the Conclusion manuscript. The work carried out in the During the last three years it has been author's laboratory has been continudemonstrated that eukaryotic cells con- ously supported by the Swiss National tain at least five DNA polymerases and Science Foundation and the Kanton of genetic experiments in yeast have indi- Zi~rich. cated that three of these might be required for replication of the [~uclear References DNA. DNA polymerase e appears to 1 Baker, T. A. and Komberg, A. (1991) BNA have functional roles both in DNA Replication (2nd edn), Freeman
repair and DNA replication. DNA polymerase ~ and DNA polymerase 6
58
2 Wang, T. S-F. (1991) Annu. Rev. Biochem. 60, 513-552 3 Burgers, P. M. J. et al. (1990) Eur. J. Biochem.
(Engler, J. A. and Wang, T. S-F., eds) SpringerVerlag (in press) 26 Yoder, B. L. and Burgers, P. M. J. (1991) J. Biol. Chem. 266, 22689-22697 27 Lee, S-H., Pan, Z-O. and Hurwitz, J. J. Biol. Chem. (in press) 28 Bollum, F. J. and Potter, V. (1957)J. Am. Chem. Soc. 79, 3603-3604 29 Johnson, L. M. et al. (1985) Cell 43, 369-377 30 Boulet, A. et al. (1989) EMBO J. 8, 1849-1854 31 Sinha, N. K., Morris, C. F. and Alberts, B. M. (1980) J. Biol. Chem. 255, 429(~4303 32 Alberts, B. M. et al. (1983) Cold Spring Harb. Symp. Quant. Biol. 47, 655-668 33 KornbergA. (1982 supplement) DNA Replication, F~eeman 34 McHenry,C. S. and Johanson,K. O. (1984) in Proteins Involved in DNA Replication, Advances in Experimental Biology and Medicine
(Hfibscher.U. and Spadari, S., eds)(Vol. 179),
pp. 315-319, Plenum 35 Ottiger, H-P. and Hfibscher, U. (1984) Prec. Natl Acad. Sci. USA 81, 3993-3997 36 McHenry, C. S. (1991) J. Biol. Chem. 266,
19127-19130
Maki, H., Maki,S. and Kornberg,A. (1988) J. Biol. Chem. 263, 6570-6578 38 Maki, S. and Komberg,A. (1988)J. Biol. Chem. 37
263, 6555-6560
39 Studwell, P. S. and O'Donnell, M. (1990) J. Biol. Chem. 265, 1171-1178 40 Stillman, B. et al. in Eukaryotic DNA Polymerases (Engler, J. A. and Wang, T. S-F.,
eds)Springer-Verlag(in press)
41 ThBmmes,P. and Hfibscher,U. (1990) Eur. J. Biochem. 194, 699-712 42 Hurwitz, J., Dean, F. B., Kwong, A. D. and Lee, S-K. (1990) J. Biol. Chem. 265, 18043-18046 43 Tsurimoto, T. and Stillman, B. (1989) EMBO J.
8, 3883-3889
Other forms of publication have been recently discussed. The launch of an FAOB research journal was originally debated over ten years ago but discussion has now broadened to include other kinds of publication, such as a review-type publication or a members' magazine. The primary objectives of the FAOB in launching a new publication are still a topic of debate: should it increase international prestige, generate income, or create a vehicle for enhancing communication and promoting a sense of identity in members? It is hoped that some consensus on this issue will soon be reached. The decision is an important one. The launch of any publication will reflect how
Constituent Members view the future role of the FAOB in the region over the next few years. Having witnessed much of the progress that has been achieved over the years, I still feel that the essential spirit of the FAOB has remained the same, namely that it is a Federation that exists for and exists through the cooperation of Constituent Members. If all those involved in the FAOB can retain this spirit of cooperation, then I am sure that the FAOB can find solutions to its immediate problems and have an even greater impact on the development of biochemistry in the AsianPacific region in the third decade of its existence.
TALKINGPOINT ALL PROKARYOTIC and eukaryotic cells contain several DNA polymerases. Besides replication of the chromosomal DNA, the cell requires DNA polymerases for other DNA synthesis events such as DNA repair, nuclear DNA recombination and replication of extrachromosomal (i.e. plasmid or mitochondrial) DNA. In bacteria, these tasks are shared by the
three different DNA polymerases I, II and III. In some instances plasmids and bacteriophages (e.g. T4 or T7) encode their own DNA polymerases ]. At least five different DNA polymerases have been identified and purified from eukaryotic cells 2. The subunit composition of the yeast enzymes is nearly identical to the one in higher eukaryotes which has lead to a new nomenclature for eukaryotic DNA polymerases 3. According to this new classification, all eukaryotic polymerases are denoted by greek letters (Table I).DNA polymerase (~ (reviewed in Ref. 4) is important in nuclear DNA replication. Recent data strongly suggest that it has a role in the initiation of the lagging strand of the replication fork 5. DNA polymerase ~ (reviewed in Ref. 6) is the major repair
DNA polymerase e: in search of a function .
.
.
.
.
.
.
The current model of eukaryotic DNA replication involves the two DNA polymerases 8 and (x as the leading and lagging strand enzymes, respectively. A DNA polymerase first discovered in yeast has now been found in all eukaryotic cells and is termed DNA polymerase E. In yeast, the gene for DNA polymerase E has recently been found to be essential for viability, raising new questions about its functions.
enzyme in the nucleus and appears to be involved in DNA recombination 7. DNA polymerase y (reviewed in Ref. 6) carries out DNA replication of the mitochondrial genome. DNA polymerase 8 (reviewed in Refs 8 and 9), first described in 1976, was originally distinguished from DNA polymerase ~ by t~ v its 3 5 proofreading exonuclease activity ]°. Subsequently, all eukaryotic 3'-~5' exonuclease-containing DNA polymerases were called DNA polymerase 8. DNA polymerase 5 is involved in nuclear DNA replication and is the Ulrich Hiil~©her and Pia Th6mmes are at the prime candidate for replication of the Department of Pharmacologyand Biochemistry, Universityof Z0rich-lrchel, leading strand of the replication fork. Winterthurerstrasse 190, CH-8057 Z0rich, On model homopolymer template/ Switzerland. primers [e.g. poly(dA)/oligo(dT)] con© 1992, ElsevierScience Publishers, (UK) 0376-5067/92/$05.00
taining long single-stranded regions, this enzyme requires an auxiliary protein, called proliferating cell nuclear antigen (PCNA), for processive DNA synthesis8. In higher eukaryotic cells, a second 3'45' exonuclease-containing DNA po]ymerase was isolated from calf thymus n and HeLa ceils ]2 which was independent of PCNA for processive DNA synthesis. A similar enzyme had been discovered more than 20 years ago in yeasP 3. Originally known as DNA polymerase II (Table ID, this enzyme is now called DNA polymerasec (reviewed in Ref. 14). Table II gives a brief history of DNA polymerases (~, 8 and ~. What is the function of DNA polymerase E? An involvement in the
55
TIBS 17 - FEBRUARY 1 9 9 2 and Table I). DNA pol~nerase ~ is com-
Table I. The five eukaryotic DNA polymerases a
DNA Yeast Biological polymerase gene function
(z
Catalytic subunit (kDa)
Other Function polypeptides (kOa)
posed of four polypeptides of approximately 160-185, 70, 60 and 50 kDa 4, and DNA polymerase 8 contains two polypeptides of 125 and 50 kDa 9. The sub-
160-185
70 50/60
40
-
unit structure of DNA polymerase E is still uncertain. Studies from yeast and calf thymus suggest that two forms exist (reviewed in Ref. 9). One form
Replicationof mitochondrial DNA POL3 Replicationof leadingstrand
125
50
Unknown
125
48
Unknown
POL2 Repairof nuclear DNA,
210-230 or 125-140
29/30/ 34/80 or 4O
All unknown
POLl
Replicationof laggingstrand, initiation of leading strand, repair ofnuclearDNA
J3
-
T
MIP1
Repairof nuclearDNA, recombination
replication of nuclear DNA
Anchor?. Primase heterodimer
(called DNA polymerase e* and, earlier,
asee Ref. 3 for a revisednomenclatureof eukaryotic DNA polymerases.Furtherreferencesare cited with-
in the text.
II in yeast) has a catalytic subunit of 125--140 kDa 23 and possibly one smaller subunit of 40 kDa TM, while the other form (called DNA polymerase ~ or earlier If* in yeast) is more complex with a catalytic subunit of 215-230 kDa and several subunits of between 30 and 70 kDa 24. The role of these two forms is not yet clear but they both exist in crude
Year
Protein
Comment
Ref.
extracts from yeast 24 and calf thymus (T. Weiser and U. HObscher, unpublished data). The sequence of yeast DNA polymerase c predicts a catalytic subunit of >250 kDa 15. Genetic experiments with the DNA polymerase c gene, however, indicated that the C-terminal half of DNA polymerase ~ was not essential for growth ~5. The N-terminal part coding for a 130 kDa DNA polymerase e was enough to restore viability of the cells, although with a slower growth rate. The catalytic core of mammalian DNA polymerase e (258 kDa) might be composed of two segments that are linked via a proteasesensitive area 23, thus resulting in either an enzyme of >200 kI)a or of 140 kDa. The 140 kDa DNA polymerase c is able to replicate a primed M13 DNA efficientlyTM. However, the mode of replication by DNA polymerase e appears to be more complex. The yeast 2s,26 and the human 27 enzymes become inactive upon addition of physiological salt concentrations. Replication activity is only restored if DNA polymerase ~ is supplemented with auxiliary proteins such as replication factor A (RF-A), PCNA and replication factor C (RF-C)2s,26. Replication of primed M13 in vitro is achieved by yeast DNA polymerase c,
1957
pol a a
Discoveryin calfthymus
28
together with those three auxiliary pro-
1970
pol ~b
Discovery in yeast
1976 1985 1986 1987 1988 1989 199o 1990
pol 6 yeast pol ac PCNA" PCNA pol E yeast pol ~
Discoveryin calf thymus Essentialgene function Auxiliaryproteinfor pol 5 Essentialfor SV40 replication in vitro PCNA-independent form Es.~entialgene function New nomenclature Esser)tialgene function
13 10 29 16 17 11 30 3 15
teins, at a speed that reaches that 0bserved in vivo2S,2E For this the 140 kDa proteolysed subunit of DNA polymerase
repair of nuclear DNA is suggested by its ability to complement in repair assays. However, that its gene is essential for yeast viability suggests an important role in DNA replication. How does this fit into our current image of the replication process?
DNA repair system yielded a iS-like DNA polymerase which was independent of PCNA~9. Based on an agreement reached at the Cold Spring Harbor Symposium in 1989, the PCNA-independent DNA polymerase 8 is now called DNA polymerase ca. For some time it was not possible to isolate DNA polDNA polymetaseE ymerases 8 and c from the same tissue, The finding that PCNA is an auxiliary Recently this has been achieved in protein of DNA polymerase 8 ~6and also HeLa cells 2°, human placenta 2~ and calf an essential factor for SV40 replication thymus 22. Biochemical comparison in vitro n brought about a renaissance of clearly distinguished DNA polymerase ~ interest in DNA polymerase 6. There from DNA polymerases 8 and (~/primase. was initial controversy, however, when The catalytic subunit of DNA polanother DNA polymerase was dis- ymerase ~ has been cloned and secovered u,18 that was independent of quenced from the yeast Saccharomyces PCNA for processive synthesis on cerevisiae. The gene was found to be model homopolymer templates such as essential for yeast viability, and a funcpoly(dA)/oligo(dT). This so-called PCNA- tion for this third essential DNA polindependent DNA polymerase 8 also ymerase in replication was therefore contained an associated proofreading suggested 15. 3'45' exonuclease activity. At about the DNA polymerases (z, 8 and ~ have difsame time, fractionation of an in vitro ferent subunit compositions (Refs 20-22
Table 2. Chronological history of the discovery of DNA polymerases ~, ~ and E
all pols
yeast pol E
a pol, DNA potymerase,
b Yeast DNA potymeraseEwas called DNA polymeraseII until 1990. c Yeast DNA polymerasea was called DNA polymeraseI until 1990. d PCNA,proliferatingcell nuclearantigen.
56
g was sufficient (P. M. J. Burgers, pers. commun.). In the SV40 in vitro replication systern ~7, DNA polymerase ~ from HeLa cells could not substitute for DNA polymerase 6. This enzyme was composed of a catalytic subunit of 210 kDa. If, however, the yeast DNA polymerase (likely a proteolysed 140 kDa E* form)
TIBS 17 - FEBRUARY1992
was used, a partial replacement of DNA polymerase 8 was observed. Since the high molecular weight DNA polymerase was originally isolated as a repair activity~9and the proteolysed DNA polymerase e* form can efficiently replicate in vitro, we suggest that proteolysis might not be merely an artifact of isolation, but might enable different functional forms to exist in the cell.
Howmightthree DNApolymerasesact at the replication fork? A model originally proposed by Sinha et a l ) 1 predicted that replication of the
leading and lagging strand might be performed by two different polymerase activities which have to be coordinated in some way. According to this model (Fig. 1), one DNA polymerase molecule replicates the leading strand continuously and the other replicates the lagging strand discontinuously. The lagging strand bends back by 180° on itself to form a loop at the replication fork. This would allow the dimeric DNA polymerase to move along both template strands in the 'same' direction without violating the 5'~3' directionality rule. When the lagging strand DNA polymerase reaches the Okazaki fragment synthesized during the previous round, the freshly synthesized DNA is threaded through the enzyme, allowing the DNA polymerase to recycle to a newly exposed single-stranded region placed 3' to the previous priming site for initiation of the next Okazaki fragment. In this way, a dimeric DNA polymerase holoenzyme would guarantee an efficient and coordinated progression of the advancing replication fork. Data supporting this model accumulated in the T4 DNA replication system 3z, for the Escherichia coli DNA polymerase III holoenzyme33,34and for various forms of the DNA polymerase holoenzyme from calf thymus3E In E. coli this dual DNA polymerase funcffOn appears to be founded on the asymmetry of the DNA polymerase III holoenzyme3E Three subunits of the DNA polymerase III core polypeptides are identical in both the leading and the lagging DNA polymerase, but each complex is associated with a variety of different additional subunits. The composition of these subunits provides each half of this twin DNA polymerase III with properties suited for replication of either the leading and the lagging strand. DNA polymerase IIl was isolated as an asymmetric dimer with 22 polypeptides 37 of which one subunit
(the x-subunit) appears to s' s' have the capacity to dimerize the DNA polymerase IIl core ~. The ~!'~ t structural basis of the .. ,~ asymmetric dimer DNA polymerase III may be due ......~ . . . _ to the occurrence of two . . ~ '.'~" s' related but differentially ~ . . processed accessory poly~ymerase~ "'-. "~.~ peptides in the same com1 ~ _ 'a°g~ng i:~ "'..... Y~_ plex (see, for example, '~ .:':~.;~ :~ , ' : " • Refs 38 and 39). ~:i~,,~6,~_,~J Work by Stillman et al. 5,4° indicated that sequential initiation of leading and lagging strand synthesis at the replication origin of SV40occurs by two different DNA polymerases, and 8. After the origin of Figure 1 The asymmetric DNA replication fork. Most organisms replication is unwound by require an asymmetrictype of DNAreplication fork for the combined action of T replicationof both DNAstrands. Dotted lines, parental antigen and the singlestrands; solid lines, newlyreplicated DNA;black bars, stranded DNA binding primers. protein (SSB, also called RF-A or RP-A), DNA polymerase a/primase initiates DNA syn- of RF-C to a reaction mixture containing thesis of the lagging strand (see Refs 41 DNA polymerase ~, PCNA and RF-A. The and 42 for a recent review on enzymes high degree of functional similarities and proteins acting at the replication between DNA polymerase 5 and ~ in fork). When a short stretch of DNA has their capacities to form a holoenzyme been formed, the two DNA polymerase structure with the same auxiliary proauxiliary proteins PCNA and RF-C come teins zS, and the fact that, in yeast, both into action by forming a tight complex enzymes are essential gene products 1s,3°, at the 3'OH end of the growing DNA suggest that both DNA polymerases chain in the presence of ATP. This bind- might function in chromosomal DNA ing prevents DNA polymerase a/pri- replication. mase from further binding to the 3'OH A prospective working model is outend and instead allows DNA polym- lined in Fig. 2. After unwinding at a erase ~i to bind and start synthesis of given origin of replication, the initiating the leading strand. The dissociated DNA polymerase (DNA polymerase DNA polymerase ~t/primase cycles to a/primase) synthesizes primers and the next priming site on the lagging short pieces of DNA for lagging strand strand to resume discontinuous repli- synthesis at both forks. The primase cation. Simultaneously, DNA pol- tightly associated with DNA polymerase ymerase ~ continues to synthesize the a makes it an indispensible enzyme for leading strand. The lagging strand syn- priming events at the lagging strand. thesized by DNA polymerase c~ serves Following this initiation of synthesis, to initiate the synthesis of the leading DNA polymerase a/primase is replaced strand, subsequently carried out by DNA by other DNA polymerases for the elonpolymerase ~iand its auxiliary proteinsE gation mode (for DNA polymerase How does DNA polymerase E, the switch, see Ref. 5). Figure 2 depicts how third essential DNA polymerase come three DNA polymerases might share the into play? DNA polymerase auxiliary replication tasks during the elongation proteins, such as RF-A, RF-C and PCNA, event: the leading strand is replicated are required by DNA polymerase ~5for by a processive DNA polymerase (DNA replication of natural DNAs,43. These polymerase ~5or DNA polymerase e). At three proteins also appear to interact the same time the initiating DNA polwith DNA polymerase E but probably in ymerase (DNA polymerase a/primase) a more complex way than with DNA continues to synthesize primers and polymerase ~i25,26. As mentioned above, short pieces of DNA at the lagging DNA replication by DNA polymerase e strand. Subsequently, the Okazaki fragbecomes resistant to salt upon addition ments are synthesized to completion by
%
57
TIBS 17 - F E B R U A R Y 1 9 9 2
Initiation
,"
... •.""
.....................................' , , . . . . .
%
". ''.. ~ % ~
~®
............ s' J I ~( ( ( [ 3'.
; - ................. 5' LI..I..J..~..~.~.'[3'
""".,° ". ....•. . .
..-"
.......... ..............................
Elongation
~
' ."
l ..................................... " r . , . . .... ...." •" "
! ~ "'"
.* •."
.." ../ •'
: ©~ "=" " C ,~
......
~~~~
"
io.~
_c'~l.......
[
~
~$ ~
:#j~
..'K~,~ "
"::'Y" "'"'"
~
I ~ .=
~
.......". ....
~:
".~:...........................
~~5,
"T':f'j'5
~
-a ~ . . ~ ~
~
~,..,"~g,~ ~" ~ ~
~ ~ . , . . ~ - - , , / ~ , ~ , ~
~';,'f
"
a" ~ ¢ ~ :~. ~~ :~ ~"
.~.....-
~ m /
-
191, 617-618 4 Lehman, I. R. and Kaguni, L. S. (1989) J. Biol. Chem. 264, 4591-4595 5 Tsurimoto, T., Melendy, T. and Stillman, B. (1990) Nature 346, 534-539 6 Fry, M. and Loeb, L. A. (1986) Animal Cell DNA Polymerases, CRC Press Inc. 7 Nowak, R., Woszczynski, M. and Siedlecki, J. A. (1990) Exp. Cell Res. 191, 51-56 8 Downey, K. M., Tan, C-K. and So, A. G. (1990) BioEssays 12, 231-236 9 Bambara, R. A. and Jessee, C. B. (1991) Biochim. Biophys. Acta 1088, 11-24 10 Bymes, J. J., Downey,K. M., Black, V. L. and So, A. G. (1976) Biochemistry 15, 2817-2815 11 Focher, F. et al. (1988) Nucl. Acids Res. 16, 6279-6295 12 Syv~oja,J. and Linn, S. (1989) J. Biol. Chem. 264, 2489-2497 13 Wintersberger, U. and Wintersberger, E. (1970) Eur. J. Biochem. 13, 11-19 14 Suv~oja, J. (1990) BioEssays 12, 533-536 15 Morrison, A. et al. (1990) Cell 62, 1143-1151 16Tan, C-K., Castillo, C., So, A. G. and Downey, K. M. (1986) J. Biol. Chem. 261, 12310-12316 17Prelich, etal. (1987)Nature326, 471-475 18 Focher, G. F. et al. (1989) Nucl. Acids Res. 17, 1805-1821 19 Nishida, C., Reinhard, P. and Linn, S. (1988) J. Biol. Chem. 263, 501-510 20 Syv~oja,J. et al. (1990) Prec. Natl Acad. Sci. USA 87, 6664-6668 21 Lee, M. Y. W. T., Jiang, Y., Thang, S-J. and Toomey, N. L. (1991) J. Biol. Chem. 266, 22 Weiser, T. et al. (1991) J. Biol. Chem. 266,
10420-10428 23 Kesti, T. and Syv,oja, J. (1991)J. Biol. Chem.
;.."
266,6336-6341 24 Hamatake, R. K. et al. (1990) J. Biol. Chem.
265, 4072-4083 25 Burgers, P. M. J. in Eukaryotic DNA Polymerases
Figure 2 Model for eukaryotic DNA replication, depicting how three DNA polymerases might participate in DNA replication. For explanation see text. Dotted lines, parental strands; solid lines, newly replicated DNA; black bars, primers.
a third DNA polymerase (DNA polymerase e or DNA polymerase ~). It is conceivable that various subsets of auxiliary protein compositions (PCNA, RF-A, RF-C and others, such as DNA hellcases) might tune the DNA polymerases and e to their respective functional roles. This model would explain how proofreading of the two strands might occur since both DNA polymerases 8 and c contain an intrinsic 3'~5' exonuclease9.
appear to form holoenzymes with the same auxiliary subunits and replicate DNA in v i t r o at physiological rates. The functional similarity of DNA polymerases and their auxiliary proteins from yeast to man strongly argues for the conservation of the eukaryotic rep]ication fork. Acknowledgements
We thank C. C. Kuenzle and M. Berchtold for critical reading of the Conclusion manuscript. The work carried out in the During the last three years it has been author's laboratory has been continudemonstrated that eukaryotic cells con- ously supported by the Swiss National tain at least five DNA polymerases and Science Foundation and the Kanton of genetic experiments in yeast have indi- Zi~rich. cated that three of these might be required for replication of the [~uclear References DNA. DNA polymerase e appears to 1 Baker, T. A. and Komberg, A. (1991) BNA have functional roles both in DNA Replication (2nd edn), Freeman
repair and DNA replication. DNA polymerase ~ and DNA polymerase 6
58
2 Wang, T. S-F. (1991) Annu. Rev. Biochem. 60, 513-552 3 Burgers, P. M. J. et al. (1990) Eur. J. Biochem.
(Engler, J. A. and Wang, T. S-F., eds) SpringerVerlag (in press) 26 Yoder, B. L. and Burgers, P. M. J. (1991) J. Biol. Chem. 266, 22689-22697 27 Lee, S-H., Pan, Z-O. and Hurwitz, J. J. Biol. Chem. (in press) 28 Bollum, F. J. and Potter, V. (1957)J. Am. Chem. Soc. 79, 3603-3604 29 Johnson, L. M. et al. (1985) Cell 43, 369-377 30 Boulet, A. et al. (1989) EMBO J. 8, 1849-1854 31 Sinha, N. K., Morris, C. F. and Alberts, B. M. (1980) J. Biol. Chem. 255, 429(~4303 32 Alberts, B. M. et al. (1983) Cold Spring Harb. Symp. Quant. Biol. 47, 655-668 33 KornbergA. (1982 supplement) DNA Replication, F~eeman 34 McHenry,C. S. and Johanson,K. O. (1984) in Proteins Involved in DNA Replication, Advances in Experimental Biology and Medicine
(Hfibscher.U. and Spadari, S., eds)(Vol. 179),
pp. 315-319, Plenum 35 Ottiger, H-P. and Hfibscher, U. (1984) Prec. Natl Acad. Sci. USA 81, 3993-3997 36 McHenry, C. S. (1991) J. Biol. Chem. 266,
19127-19130
Maki, H., Maki,S. and Kornberg,A. (1988) J. Biol. Chem. 263, 6570-6578 38 Maki, S. and Komberg,A. (1988)J. Biol. Chem. 37
263, 6555-6560
39 Studwell, P. S. and O'Donnell, M. (1990) J. Biol. Chem. 265, 1171-1178 40 Stillman, B. et al. in Eukaryotic DNA Polymerases (Engler, J. A. and Wang, T. S-F.,
eds)Springer-Verlag(in press)
41 ThBmmes,P. and Hfibscher,U. (1990) Eur. J. Biochem. 194, 699-712 42 Hurwitz, J., Dean, F. B., Kwong, A. D. and Lee, S-K. (1990) J. Biol. Chem. 265, 18043-18046 43 Tsurimoto, T. and Stillman, B. (1989) EMBO J.
8, 3883-3889