ABSTRACT
We used the known sequence of the
Saccharomyces cerevisiae
DNA polymerase
[gamma]
to clone the genes or cDNAs encoding this enzyme in two other yeasts,
Pychia pastoris
and
Schizosaccharomyces pombe
, and one higher eukaryote,
Xenopus laevis
. To confirm the identity of the final
X.laevis
clone, two antisera raised against peptide sequences were shown to react with
DNA polymerase
[gamma]
purified from
X.laevis
oocyte mitochondria. A developmentally regulated 4.6 kb mRNA is recognized on
Northern blots of oocyte RNA using the
X.laevis
cDNA. Comparison of the four DNA polymerase
[gamma]
gene sequences revealed several highly conserved sequence blocks, comprising an
N-terminal 3
' ->
5
'
exonuclease domain and a C-terminal polymerase active center interspersed with
[gamma]
-specific gene sequences. The consensus sequences for the DNA polymerase
[gamma]
exonuclease and polymerase domains show extensive sequence similarity to DNA
polymerase I from
Escherichia coli
. Sequence conservation is greatest for residues located near the active centers
of the exo and pol domains of the
E.coli
DNA polymerase I structure. The domain separating the exonuclease and
polymerase active sites is larger in DNA polymerase
[gamma]
than in other members of family A (DNA polymerase I-like) polymerases. The
S.cerevisiae
DNA polymerase
[gamma]
is atypical in that it includes a 240 residue C-terminal extension that is not found in the other members of the DNA
polymerase [gamma] family, or in other family A DNA polymerases.
Our laboratory is interested in the structure and function of DNA polymerase [gamma], the enzyme responsible for replication of mitochondrial DNA. In this
paper we report the molecular cloning of
Xenopus laevis
DNA polymerase [gamma] cDNA using the sequence of the gene encoding DNA polymerase [gamma] in the budding yeast
Saccharomyces cerevisiae
(
1
). Although
S.cerevisiae
and
X.laevis
are highly divergent species, the purified
X.laevis
DNA polymerase [gamma] (
2
,
3
) shares a number of features with the yeast enzyme. Both enzymes efficiently
utilize poly(rA)[middot]oligo(dT) templates, respond similarly to polymerase inhibitors, possess
associated 3' -> 5' exonuclease activity and contain a DNA polymerase catalytic
subunit of ~140 kDa. These features are shared with DNA polymerase [gamma] purified from other sources as well (
4
-
8
), although the catalytic subunit of the
Drosophila
enzyme is somewhat smaller than that for vertebrates.
The primary sequence of
S.cerevisiae
DNA polymerase [gamma] (
1
) shows similarity to those of other DNA polymerase I-like (family A) DNA polymerases in the polymerase and exonuclease domains
(
9
,
10
). However, the
S.cerevisiae
DNA polymerase [gamma] presents a rather poor match to the overall consensus sequence for
family A polymerases in that it contains only one third of the highly conserved
residues identified by Braithwaite and Ito (
11
). We first designed PCR primers using sequences in the pol domain to obtain the
sequence of the gene from a second budding yeast,
Pychia pastoris
. This sequence information permitted the design of new primers to clone DNA
polymerase [gamma] from the fission yeast
Schizosaccharomyces pombe
. The homology among these three fungal DNA polymerase [gamma] sequences enabled us to clone the cDNA for the
X.laevis
enzyme. In this paper we report these novel gene sequences and provide a
comparison among the [gamma] subfamily of DNA polymerases and other family A DNA polymerases.
Kluyveromyces lactis
genomic DNA and
Pychia pastoris
strain SB 1099 were obtained from N. Dean (SUNY, Stony Brook, NY) and
Schizosaccharomyces pombe
strain SP1301 was obtained from R. Sternglanz (SUNY, Stony Brook, NY). The
cells were either grown in YPAD broth (2% peptone, 1% yeast extract, 40 mg/ml
adenine sulfate and 2% glucose) or YE (2.5% ethanol, 2% yeast extract) solid
medium containing 1.5% agar.
The PCR amplification reactions were performed using degenerate oligonucleotide
primers to amplify sequences from 0.1 [mu]g genomic DNA isolated from
P.pastoris
,
K.lactis
or
S.pombe
according to Moreno
et al
. (
12
). At sites of 2- or 3-fold degeneracy, mixed nucleotide residues were incorporated. At
sites of 4-fold degeneracy, inosine was incorporated. For PCR with
P.pastoris
genomic DNA, amplification reactions were performed for 30 cycles of 94oC for 1 min, 50oC for 1 min and 72oC for 2 min under reaction conditions suggested by the
manufacturer of
Taq
polymerase (Perkin Elmer-Cetus). The amplification reactions using
S.pombe
DNA were performed for three cycles of 94oC for 1 min, 37oC for 1 min and elongation at 72oC for 2 min, followed by an additional 30 cycles of 94oC for 1 min, 45oC for 1 min and 72oC for 2 min. Fragments of the expected size were
gel purified and cloned into the pCR
II vector (Invitrogen). Cloned PCR fragments were labeled by random priming and
used to screen either a partial genomic
P.pastoris
library or [lambda]Zap cDNA libraries (
S.pombe
library kindly provided by D. Beach;
X.laevis
library kindly provided by M. Roth). Hybridizations were continued for 16 h at
65oC in 5* SSC (1* SSC = 0.15 M NaCl, 0.015 M Na
3
citrate), 5* Denhardt's solution, 0.5% SDS and 100 [mu]g/ml calf thymus DNA. A final wash was performed in 1* SSC, 0.1% SDS at 65oC for 1 h.
A fragment of the
X.laevis
DNA polymerase [gamma] cDNA was obtained by two cycles of nested PCR. In the first round the
primers employed the sense strand of codons AVENTWLT and the antisense strand
of codons WTRA(F+M)FCQ. The first round amplification reactions were performed
for three cycles of 94oC for 1 min, 35oC for 1 min and 72oC for 2 min, followed by an additional 30 cycles of 94oC for 1 min, 45oC for 1 min and 72oC for 2 min. For the second round of amplification
the degenerate primers represented the sense strand of codons FVGADVD and the
antisense strand of codons AK(V+I)FNYGR, using the product of the first round
reaction as the template.
The 5'-end of the
X.laevis
DNA polymerase [gamma] was obtained by 5'-RACE (
13
) using
Pfu
polymerase (Stratagene) to reduce the potential for replication errors.
Positive clones were sequenced using Sequenase (United States Biochemical
Corp.) in the chain terminator (dideoxy) method using gene-specific primers.
Total RNA was prepared from defined numbers of staged
X.laevis
oocytes as described (
14
). RNA was resuspended in 10 mM Tris, pH 8, 1 mM EDTA, 0.5% SDS at a
concentration of 1 [mu]l/oocyte equivalent for oocyte stages I-III and 2 [mu]l/oocyte for oocyte stages IV-VI. RNA was subjected to electrophoresis on a formaldehyde-agarose gel (
15
) and transferred to a Nytran membrane (Schleicher & Schuell). Hybridization with a probe prepared by random priming (
16
) of a 3.6 kb restriction fragment containing the entire coding region of the
X.laevis
DNA polymerase [gamma] cDNA, excluding the signal sequence, was performed at 65oC in 6* SSC, 5* Denhardt's reagent, 100 [mu]g/ml calf thymus DNA, 0.5% SDS. Filters were
washed with 2* SSC, 0.5% SDS followed by 0.1* SSC/0.1% SDS at 65oC and hybridization was detected by autoradiography.
Peptides corresponding to two conserved regions of the
X.laevis
DNA polymerase [gamma] sequence, TRRAVEPTWLTASNC and KVFNYGRIYGAGC, were synthesized by
Biosynthesis Inc., coupled to ovalbumin and injected into rabbits (nos 338 and
339 respectively) to prepare antisera using standard techniques (
17
). Antipeptide antisera were prepared by Cocalico Biologicals. A separate
polyclonal antibody raised against purified 140 kDa
X.laevis
DNA polymerase [gamma] by N. Insdorf of this laboratory was used as a positive control. These
antisera and the pre-immune sera for the two antipeptide reagents were diluted 1:20 000 for use
in immunoblotting experiments. Following overnight incubation with the primary
antisera, filters were washed with phosphate-buffered saline containing 0.5% Tween 20 and incubated with a 1:5000
dilution of calf alkaline phosphatase-conjugated goat anti-rabbit antibodies (KPL Laboratories). Following additional washing
of the membrane, proteins were identified with colorimetric detection reagents
(KPL Laboratories).
In our first attempts to use the sequences of
S.cerevisiae
DNA polymerase [gamma] to clone a second DNA polymerase [gamma] we performed PCR reactions with genomic DNA derived from both
K.lactis
and
P.pastoris.
Two amino acid sequences conserved among the polymerase domains of the
S.cerevisiae
DNA polymerase [gamma] gene and polymerase 1-like DNA polymerase genes, FNYGRIY and SIHDEIR, were used to design
degenerate oligonucleotides. Fragments of the expected size were obtained from
both amplification reactions. Since the
K.lactis
sequence was found to be very closely related to that of the
S.cerevisiae
DNA polymerase [gamma], further analysis of this gene was not pursued. The
P.pastoris
PCR product was used to isolate a clone containing a 4 kb
Eco
RI insert of genomic DNA. This insert was completely sequenced. The sequence was
found to lack both an initiator methionine and homology to the extreme N-terminus of
S.cerevisiae
DNA polymerase [gamma]. An overlapping clone was isolated and the N-terminus of DNA polymerase [gamma] was sequenced. The composite sequence of the
P.pastoris
DNA polymerase [gamma] gene has been submitted to GenBank (accession no. U49510).
Four degenerate oligonucleotide primers derived from conserved polymerase domain
sequences of the three yeast DNA polymerase [gamma] genes were used to amplify fragments of
X.laevis
cDNA by two rounds of nested PCR as described in Materials and Methods. This
provided a 195 bp product that was cloned into the pCR II vector and sequenced.
The sequence showed 60-65% amino acid identity to each of the sequences of the yeast DNA
polymerase [gamma] genes. This fragment was used as a probe for screening 800 000
recombinant phage plaques from an
X.laevis
ovary cDNA library (gift of M. Roth, Hutchinson Cancer Center, Seattle, WA).
Hybridizations were performed as described above. Eleven positive [lambda] clones were isolated and plaque purified and the DNA inserts with the
vector sequence carried by the [lambda] clones were recovered by
in vivo
excision. One cDNA clone was fully sequenced using gene-specific primers. This clone contained an open reading frame of 663 amino
acids with clear homology to the C-terminus of the yeast DNA polymerase [gamma] genes. This was considered to be an incomplete cDNA clone, since
it lacked an N-terminal methionine and sequences related to the N-terminal half of yeast DNA polymerase [gamma]. The same 5'-end was observed in four independent [lambda] isolates from this library, while other
isolates were even more severely truncated. PCR reactions showed that longer
isolates were not available in this library or in three other
X.laevis
oocyte and embryo cDNA libraries.
5'-RACE using
Pfu
polymerase was used to complete the cloning of the
X.laevis
DNA polymerase [gamma] (
13
). The composite sequence of the overlapping clones contains 1200 amino acids
with a predicted molecular mass of 137.3 kDa, in excellent agreement with the
estimated molecular mass of the purified catalytic subunit (
2
). The DNA sequence (data not shown; deposited in GenBank as accession no.
U49509) includes 212 residues of putative 5' untranslated sequence, including two stop codons in-frame with the initiator ATG. The extreme N-terminus represents a potential mitochondrial transit
sequence. These features suggest that the sequence contains the full extent of
the open reading frame encoding DNA polymerase [gamma]. A comparison of the protein sequence with those of other members of the
polymerase [gamma] family is shown in Figure
1
and discussed in more detail below.
To confirm that the isolated
X.laevis
cDNA encodes DNA polymerase [gamma] antisera were raised against two conserved peptide sequences from the
X.laevis
cDNA sequence as described in Materials and Methods. These two antisera, 338
and 339, were used to probe a Western blot containing DNA polymerase [gamma] purified from oocyte mitochondria. Both antisera recognized the same ~140 kDa DNA polymerase [gamma] polypeptide identified by an antisera raised against the
intact
X.laevis
polypeptide (Fig.
2
). The blots shown in Figure
2
indicate that the pre-immune serum for 339 appeared to show a relatively high background of
reactivity. All three antisera also recognized a polypeptide of ~100 kDa that appears to be a proteolytic fragment of the catalytic subunit.
These results support the conclusion that the cloned gene does encode
X.laevis
DNA polymerase [gamma].
RNA blot hybridization was performed to confirm the size and expression of the
X.laevis
DNA polymerase [gamma]. Total RNA was isolated from defined numbers of oocytes and quantities
of RNA equivalent to the contents of 10 immature oocytes or five stage IV-VI oocytes were subjected to electrophoresis on a formaldehyde-agarose gel. Lower numbers of cells were used for the mature
oocytes due to the large quantity of rRNA accumulated at these stages. The RNA
was transferred to a membrane and hybridized to a radioactive DNA probe
containing the entire coding region of the DNA polymerase [gamma] except for the signal sequence. An mRNA of ~4.6 kb was identified (Fig.
3
), in excellent agreement with the 4396 nt cDNA sequence, which does not include
a poly(A) tail. In this and other RNA blot experiments hybridization was most
intense with stage II oocytes. This result is consistent with other experiments
we have performed using internally standardized RT-PCR methods that show that
absolute mRNA levels are decreased several-fold in mature oocytes (J. Carrodeguas, unpublished observations).
We used the sequence of
S.cerevisiae
DNA polymerase [gamma] to clone the homologous genes or cDNAs from two other yeasts. The
resulting sequence relationships permitted cloning and sequencing of the cDNA
for
X.laevis
DNA polymerase [gamma]. Confirmation that the final eukaryotic gene does indeed encode DNA
polymerase [gamma] was provided by the finding that two antisera raised against conserved
sequences reacted with the purified
X.laevis
DNA polymerase [gamma] (Fig.
2
). Hybridization of the cDNA to total RNA isolated from staged oocytes revealed
a 4.6 kb mRNA that declines in abundance as oocytes progress through
vitellogenesis (Fig.
3
). This result is also consistent with the well-established pattern of accumulation of mtDNA during oogenesis in
X.laevis
. Most of the ~100 000-fold amplification of mtDNA that occurs during oogenesis is complete
by stage IV (
19
) and mtDNA replication is relatively quiescent in mature oocytes and early
embryos (
20
). A more complete analysis of the levels of DNA polymerase [gamma] protein and mRNA levels throughout oogenesis and embryogenesis is in
progress and will be presented elsewhere.
The sequence homology among members of the DNA polymerase [gamma] subfamily is rather low. Pairwise comparison of sequences typically
shows only 40-60% identity. However, alignment of all four genes (Fig.
1
) reveals an ordered pattern of conserved sequences, as diagramed in Figure
4
. All four polymerases contain N-terminal sequences recognizable as mitochondrial targeting signals,
despite the fact that the primary sequence of these signals is not conserved.
Some of these conserved sequence blocks comprise the 3' -> 5' exonuclease and polymerase active centers, conserved in
family A DNA polymerases, while others appear to be specific to the [gamma] subfamily. The observation that both polymerase and 3' -> 5' exonuclease domains are conserved in a vertebrate
DNA polymerase [gamma] suggests that the report that the 3' -> 5' exonuclease domain of porcine DNA polymerase [gamma] may reside in a separate subunit (
8
) is unlikely to be correct. It is quite possible that an N-terminal proteolytic fragment of DNA polymerase [gamma] that retained exonuclease activity may have been present in the
enzyme preparation used by Longley and Mosbaugh (
8
). Figures
1
and
4
show that the
X.laevis
and
S.cerevisiae
DNA polymerase [gamma] sequences are appreciably longer than those from
P.pastoris
or
S.pombe
. The larger size of the
S.cerevisiae
enzyme was documented in the original cloning (
1
), while the larger size of the
X.laevis
DNA sequence is consistent with the previous report of the size of its
catalytic subunit (
2
). The other two polymerases have not been studied previously. It is interesting
to note that different patterns of sequence insertions or additions create the
larger sizes of the
X.laevis
and
S.cerevisiae
polymerases. The
S.cerevisiae
enzyme contains a long C-terminal tail not found in the other DNA polymerase [gamma] sequences, or in other members of family A DNA polymerases (
11
). The larger size of the
X.laevis
enzyme results mainly from a series of at least six substantial insertions
(depending on the detailed sequence alignment) between the exo and pol domains.
We thank D. Beach, N. Dean, F. Foury, M. Roth and R. Sternglanz for reagents
used in this work and P.Fisher for review of the manuscript. This research was
supported by grant NIH GM 29681.
*To whom correspondence should be addressed
+
Present address: Department of Biology, Massachusetts Institute of Technology,
Cambridge, MA, USA
REFERENCES
Return