RESEARCH ARTICLE
DmWRNexo is a 30–50 exonuclease: phenotypic
and biochemical characterization of mutants
of the Drosophila orthologue of human
WRN exonuclease
Ivan Boubriak Æ Penelope A. Mason Æ David J. Clancy Æ Joel Dockray Æ
Robert D. C. Saunders Æ Lynne S. Cox
Received: 3 September 2008 / Accepted: 25 September 2008

Springer Science+Business Media B.V. 2008
Abstract The premature human ageing Werner’s
syndrome is caused by loss or mutation of the WRN
helicase/exonuclease. We have recently identified
the orthologue of the WRN exonuclease in flies,
DmWRNexo, encoded by the CG7670 locus, and
showed very high levels of mitotic recombination in a
hypomorphic PiggyBac insertional mutant. Here, we
report a novel allele of CG7670, with a point
mutation resulting in the change of the conserved
aspartate (229) to valine. Flies bearing this mutation
show levels of mitotic recombination 20-fold higher
than wild type. Molecular modelling suggests that
D229 lies towards the outside of the molecule distant
from the nuclease active site. We have produced
recombinant protein of the D229V mutant, assayed
its nuclease activity in vitro, and compared activity
with that of wild type DmWRNexo and a D162A
E164A double active site mutant we have created.
We show for the first time that DmWRNexo has 30–50
exonuclease activity and that mutation within the
presumptive active site disrupts exonuclease activity.
Furthermore, we show that the D229V mutant has
very limited exonuclease activity in vitro. Using
Drosophila, we can therefore analyse WRN exonu-
clease from enzyme activity in vitro through to fly
phenotype, and show that loss of exonuclease activity
contributes to genome instability.
Keywords Werner syndrome
WRN

Exonuclease
CG7670
Aging
Ageing

Recombination
Drosophila
DmWRNexo
Abbreviations
BSA Bovine serum albumin
DTT Dithiothreitol
ECL Enhanced chemiluminescence
HRP Horse radish peroxidase
IPTG Isopropyl b-D-1-thiogalactopyranoside
PBS Phosphate buffered saline
TILLING Targeting Induced Local Lesions IN
Genomes
WS Werner syndrome
WRN Protein mutated in Werner syndrome
Introduction
Cellular senescence was first described in cultured
cells by Hayflick (1965). It is thought to arise as a
consequence of various stress signals feeding into a
signalling network that results in loss of proliferative
capacity (reviewed in Cox and Faragher 2007). One
important signal for onset of senescence is DNA
damage, arising either from exposure to exogenous
I. Boubriak
P. A. Mason
J. Dockray
L. S. Cox (&)
Department of Biochemistry, University of Oxford,
South Parks Road, Oxford OX1 3QU, UK
e-mail: lynne.cox@bioch.ox.ac.uk
D. J. Clancy
R. D. C. Saunders
Department of Life Sciences, The Open University,
Milton Keynes MK7 6AA, UK
123
Biogerontology
DOI 10.1007/s10522-008-9181-3

genotoxic agents, or from endogenous sources such
as telomere attrition or the presence of stalled
replication forks (Cox and Faragher 2007). The
WRN helicase/exonuclease is important in replication
fork progression (Rodriguez-Lopez et al. 2002;
Sidorova et al. 2008), probably acting to prevent
the accumulation of recombinogenic intermediates at
stalled forks (Rodriguez-Lopez et al. 2007) or to
repair damage caused by replication fork arrest
(Dhillon et al. 2007). WRN is indeed implicated in
DNA recombination and DNA repair (reviewed by
Cheng et al. 2007; Cox and Faragher 2007). WRN
thus plays an important role in maintaining genome
stability, and its loss contributes to premature cellular
senescence (Salk et al. 1981; Faragher et al. 1993)
and organismal ageing in the progeroid Werner
syndrome (WS) (Yu et al. 1996, 1997). The impor-
tance of stress signalling in premature senescence in
Werner’s syndrome is evident from the finding that
inhibition of the stress kinase p38MAPK restores
normal proliferation characteristics to Werner syn-
drome fibroblasts (Davis et al. 2005).
In vertebrates, WRN possesses both exonuclease
and helicase domains within the same polypeptide,
whilst in E. coli, yeast and Arabidopsis, these
activities are encoded on distinct polypeptides (e.g.
RecQ helicase and RecJ nuclease in E. coli, Courcelle
and Hanawalt 1999). In Drosophila, we have recently
identified an orthologue of the exonuclease compo-
nent of human WRN, DmWRNexo (Cox et al. 2007;
Saunders et al. 2008), which allows us to investigate
the role of the exonuclease component of WRN
distinct from the helicase component, in a metazoan
animal with distinct developmental stages.
DmWRNexo is encoded by the CG7670 locus. A
strong hypomorphic piggyBac insertional allele,
CG7670e04496, results in very high levels of mitotic
recombination, as observed using the recessive
marker mwh in the adult fly wing blade (Saunders
et al. 2008), suggesting that WRN exonuclease
activity is important to maintain genomic stability.
In order to investigate the effect on flies of subtle
alterations in DmWRNexo activity, we have
employed the technique of Targeting Induced Local
Lesions IN Genomes (TILLING) (McCallum et al.
2000; Till et al. 2003) to identify a mutant allele of
CG7670, bearing a point mutation in which the
conserved aspartate 229 is altered to valine (D229V).
We have analysed genome stability in CG7670e04496/
CG7670D229V flies and show moderately elevated
levels of mitotic recombination. To investigate the
molecular basis of this increased instability, we have
also generated a recombinant version of DmWRNexo
bearing the D229V mutation, and we compare its
nuclease activity in vitro with both wild type
recombinant DmWRNexo and a nuclease-dead
mutant. This study allows us for the first time to
relate the in vivo phenotype of flies mutant for WRN
exonuclease to the in vitro biochemical activity of
DmWRNexo.
Methods
Fly stocks and TILLING
Unless otherwise noted, fly stocks were obtained from
the Bloomington Drosophila Stock Center (http://
flystocks.bio.indiana.edu/), and were maintained on an
oatmeal, yeast, cornmeal, dextrose, and agar medium
at a constant temperature of 25C. CG7670e04496 bears
a piggyBac insertion within the 50 UTR, and is a strong
hypomorph (Cox et al. 2007; Saunders et al. 2008).
The CG7670D229V allele was identified through the
Fly-TILL project (http://tilling.fhcrc.org:9366/fly/
Welcome_to_Fly-TILL.html) (McCallum et al. 2000;
Till et al. 2003). To analyze recombination as revealed
through the mwh marker, adult wing blades were dis-
sected from flies stored in 70% ethanol, mounted in
Gary’s Magic Mountant (Lawrence et al. 1986) and
analysed by brightfield microscopy (Saunders et al.
2008).
Bioinformatics and molecular modelling
DmWRNexo structure was modelled by fitting to the
PDB co-ordinates published for human WRN (2fby)
(Perry et al. 2006) using Swiss Model (Peitsch et al.
1996; Schwede et al. 2003) then imported into PyMol
v 0.99 (Delano Scientific). Residues were coloured
and the image merged with human WRN exonuclease
domain using PyMol software.
DNA preparation and site-directed mutagenesis
CG7670 DNA (Saunders et al. 2008) was subcloned
into pIVEX2.3d (Roche) using restriction sites NdeI
and SmaI, transformed into chemically competent
Biogerontology
123

E. coli DH5a and DNA prepared by standard methods
(Qiagen). Site directed mutagenesis to generate the
D229V mutation was performed using the Stratagene
Quikchange II kit according to manufacturer’s instruc-
tions, with primers: forward 50-GAAAGCTGGC
ACGTGTTTTCCCCGAGGTTAC-30 and reverse
50-GTAACCTCGGGGAAAACACGTGCCAGCTTT
C-30, under the following PCR conditions: 95C 30 s
(1 cycle) then 16 cycles of 95C 30 s, 55C 60 s, 68C
6 min 20 s, followed by digestion with DpnI for
90 min at 37C and transformation into E. coli XL1-
blue. Transformed bacteria were plated onto LB agar
supplemented with 50 lg/ml ampicillin and grown
inverted overnight at 37C. DNA sequence analysis
was performed on an ABI 3730xl DNA Analyser-
Titania platform (GeneService, Department of
Biochemistry, University of Oxford).
Recombinant protein expression and purification
For protein expression, the pIVEX2.3d vector with
appropriate insert (or empty vector for negative
controls) was transformed by heat shock into chem-
ically competent E. coli BL21 T7 Iq LysY (NEB),
grown in SOC medium for 30 min at 37C 220 rpm
then plated onto LB agar supplemented with ampi-
cillin to 50 lg/ml and incubated overnight. Individual
colonies were picked into 5 ml LB-amp and grown
with shaking for 8 h at 37C, then diluted 1:100 in
fresh LB-amp and grown overnight at 37C with
shaking. Aliquots of this overnight culture were
diluted 1:60 in a total volume of 1.8 litres of LB-amp,
grown with shaking at 37C for 3 h, induced by
addition of 0.75 mM IPTG for 3.5 h, and then
bacteria were harvested by centrifugation at
5,000 rpm 25 min 4C in a Beckman JLA10 rotor.
Medium was removed and the cell pellet frozen
overnight at -80C. Cells were lysed in buffer HS
(250 mM NaCl, 5 mM DTT, 1 mM benzamidine,
0.5 mg/ml lysozyme) for 20 min on ice and super-
natant harvested after centrifugation at 14,000 rpm
4C in a Beckman GS-15R centrifuge rotor F3602.
Bacterial lysate was adjusted to 20 mM imidazole
then applied to a 1 ml His-Trap column (GE Health-
care) and eluted over a step gradient of 40, 60, 100,
300 and 500 mM imidazole in PBS according to
manufacturer’s instructions (GE Healthcare). Column
fractions were analysed by dot blotting. Briefly, 2 ll
of each column fraction was spotted onto
nitrocellulose, air dried then the membrane was
blocked for 1 h room temperature with 5% non-fat
milk in PBS with 0.2% Tween 20, washed and probed
with HRP-conjugated anti-his monoclonal antibody
(Roche) diluted 1:500 in 0.2% milk–PBS–Tween for
1 h at room temperature, then washed and the blot
developed using the ECL system with exposure to
Hyperfilm (GE Healthcare). Fractions were also
analyzed on SDS-PAGE gels stained with Coomassie
or silver, and by Western blotting using HRP-anti-his
antibody (Roche). To confirm protein identity, major
bands from peak fractions were excised from gels and
subjected to N terminal sequencing (courtesy of Tony
Willis, Protein sequencing facility, Department of
Biochemistry, University of Oxford).
Nuclease assays
Recombinant proteins were desalted by passing
through Sephadex G25 columns into buffer DS
(150 mM Tris–HCl pH 8.0, 50 mM NaCl, 5 mM 2-
mercaptoethanol, 0.1 9 Complete EDTA-free prote-
ase inhibitors (Roche)), with the addition of glycerol
to 20% before aliquotting and storage at -80C.
PAGE-purified fluorescein 50 end-labelled 50-mer
oligonucleotide (50-FLO in Table 1) and various
complementary unlabelled oligonucleotides were
obtained from Invitrogen and stored in 10 mM
Tris–HCl pH 7.6 at 1 nmol/ll at -20C and in the
dark (see Table 1 for sequences). Duplex substrates
were annealed as 1:1.2 (molar ratio) of fluorescent
oligonucleotide: complementary oligonucleotide in
1 9 TE with 50 mM NaCl, heated to 95C for 3 min
then allowed to cool slowly to below 30C. Sub-
strates were checked for annealing using non-
denaturing SDS–PAGE. Working aliquots of
annealed substrates were kept at 4C.
Exonuclease assays were conducted using 1 pmol
of recombinant DmWRNexo protein with 20 pmol
oligonucleotide in WRN exo buffer [40 mM Tris–
HCl, pH 8.0, 4 mM MgCl2, 5 mM dithiothreitol,
0.1 mg/ml BSA (Opresko et al. 2001)] and incubated
at 18, 25 or 37C for up to 40 min. Reactions were
stopped by the addition of formamide buffer [80%
formamide, 0.5 9 TBE ± bromophenol blue and
xylene cyanol (Opresko et al. 2001)]. Products were
resolved on 14% acrylamide gels containing 8 M
urea at 200 V for 150–200 min. DNA fragments were
visualized on a Fuji FLA-3000 with filter 470/520
Biogerontology
123

and the most sensitive settings i.e. 16-bit imaging,
F-1,000 and a resolution of 50. Quantitation utilised
ImageJ (http://rsbweb.nih.gov/ij/index.html).
Results
We have recently shown that very low expression of
DmWRNexo through piggyBac insertion in the 50
UTR of CG7670 leads to very high levels of DNA
recombination, as assayed by the recessive multiple
wing hairs marker (mwh) (Saunders et al. 2008).
While cells heterozygous for mwh show the usual one
hair per wing blade cell, cells that are either
hemizygous or homozygous for mwh display tufts
of wing hairs. To investigate what, if any, impact a
milder mutation in DmWRNexo might have in
genome stability, a fly stock bearing a point mutation
altering aspartate 229 to valine was obtained by
TILLING (see ‘‘Methods’’).
Recombination in CG7670D229V flies
To analyse if the CG7670D229V mutant allele has an
impact on genomic stability, compound heterozygote
flies with one copy of the D229V allele and one copy
of the piggyBac allele (i.e. mwh1 CG7670e04496/
CG7670D229V) were scored for frequency and size of
wing blade clones. Recombination rates were com-
pared with those for flies homozygous for the
piggyBac insertional allele (mwh1 CG7670e04496/
CG7670e04496) and flies hemizygous for
CG7670e04496 (mwh1 CG7670e04496/Df(3R)Exel6178).
Interestingly, CG7670e04496 hemizygotes (Fig. 1a)
show approximately twice the rate of recombination
(for single cell clones) as observed for CG7670e04496
homozygotes (Fig. 1b), strongly suggesting that this
allele is hypomorphic, and demonstrating a gene
dosage effect. The phenotypes observed (see also Cox
et al. 2007; Saunders et al. 2008) are therefore likely
to be due to a limiting amount of DmWRNexo.
By contrast, rates of recombination in mwh1
CG7670e04496/CG7670D229V flies were much lower
(Fig. 1c), though *20-fold more single cell clones
were observed than in mwh1 CG7670e04496/CG7670?
heterozygotes (0.2 single cell clones per fly, data not
shown, see Saunders et al. 2008). Thus mutation of
aspartate D229 to valine has a marked deleterious
impact on the activity of DmWRNexo in restraining
excessive mitotic recombination in flies. In terms of
the recombination phenotype (Fig. 1d), we can derive
Table 1 Oligonucleotides used in this study
(a) Name Sequence 50–30
50-FLO dGAACTATGGCTCTCGAGTGCTAGGACATGTCTGACTACGTACAAGTCACC
B-50 GGTGACTTGTACGTAGTCAGACATGTCCTAGCACTCGAGAGCCATAGTTC
RO-65-50 (T)15GGTGACTTGTACGTAGTCAGACATGTCCTAGCACTCGAGAGCCATAGTTC
(b) Name Oligos used Substrate characteristics
ssFLO 50-FLO
50
Blunt 50-FLO ? B-50
50
50 overhang 50-FLO ? RO-65-50
50 15
(a) Sequences of oligonucleotides. (b). Substrates for nuclease assays derived from the oligonucleotides shown in part (a). Filled
circle denotes fluorescein label. Size in nucleotides is given below the substrate
Biogerontology
123

the allelic series: deficiency [ CG7670e04496 [
CG7670D229V. Note that even the relatively mild
recombination phenotype seen in mwh1 CG7670e04496/
CG7670D229V flies represents recombination rates
more than 20-fold higher than levels observed in
control flies, with a *400-fold increase in recombi-
nation frequency in the CG7670e04496 hemizygous
flies over control flies.
Molecular modelling of the D229V mutation
of DmWRNexo
To attempt to understand how the D229V mutation of
DmWRNexo may impact on recombination and
thereby result in the hyper-recombination phenotype,
we modelled its position within the protein. A
structural model of DmWRNexo was generated using
Swiss Model and PyMol (coloured green in Fig. 2a,
see also Saunders et al. 2008) according to the PDB
co-ordinates of human WRN exonuclease domain
(2fby, Perry et al. 2006, Fig. 2b). From this molecular
modelling, it can be seen that the D229 residue
(coloured yellow in DmWRNexo Fig. 2a, equivalent
to D150 in human WRN, Fig. 2b) is not one of the
acidic amino acids thought to be important for
catalysis (D162, E164 and D222 in DmWRNexo,
equivalent to D82, E84 and D143 in human WRN
and coloured blue, magenta and red in Fig. 2,
respectively), but instead lies spatially distant from
the presumptive active site of the enzyme. The close
fit of the model for DmWRNexo with respect to
human WRN is shown in the merged structure
(Fig. 2c).
Production of recombinant DmWRNexo protein
In order to investigate the molecular basis of the
hyper-recombination detected in CG7670e04496/
CG7670D229V mutant flies, we generated the D229V
mutant by site-directed mutagenesis of the wild-type
clone (Saunders et al. 2008) in the vector pIVEX2.3d.
We expressed a his-tagged recombinant version of
the mutant DmWRNexoD229V in E. coli, in parallel
Fig. 1 Recombination frequencies in wing blades of flies
bearing mutations in the gene for DmWRNexo. Recombination
is determined by appearance of the multiple wing hair marker
(observed as tufts of hairs in wing cells instead of the usual
single wing blade hair per cell) and the frequency of such
clones is plotted against the number of cell cycles since the
recombination event, as determined from number of cells
within the clone. Note that all flies are originally heterozygous
for the recessive mwh marker: the clones of cells with scorable
tufts of hair (i.e. homozygous for mwh1) arise via mitotic
recombination. a CG7670e04496 hemizygotes mwh1
CG7670e04496/Df (3R)Exel6178, b CG7670e04496 homozygotes
mwh1 CG7670e04496/CG7670e04496, c CG7670D229V/mwh1
CG7670e04496. Note the different scales of the y axes in parts
(a), (b) and (c). d Data from parts (a), (b) and (c) are combined
to show the allelic series for severity of recombination
phenotype
b
Biogerontology
123

with wild type DmWRNexo and a double mutant
form DmWRNexoD162A,E164A in which presumptive
active site residues (D162 and E164) are mutated to
alanine, that we predict (see Fig. 2) should lack
exonuclease activity. These proteins were purified by
metal ion affinity chromatography. Peak fractions of
his-tagged protein were found to elute between 300
and 500 mM imidazole from dot blot analysis (data
not shown) and Coomassie-stained gels (Fig. 3a). For
each peak fraction, following desalting, a major band
of *55 kDa representing purified recombinant
DmWRNexo was observed on silver-stained SDS–
PAGE (Fig. 3b). Western blotting was also per-
formed using an antibody against hexa-histidine,
confirming that the major Coomassie-stained
DmWRNexo recombinant protein bands did indeed
possess the expected histidine tag (data not shown).
The identity of DmWRNexo was further confirmed
by Edman degradation sequencing of excised bands;
in addition to the major band for wt DmWRNexo, a
Fig. 3 DmWRNexo recombinant protein. a Coomassie-
stained protein gels of peak fractions eluting between 300
and 500 mM imidazole for all purified recombinant proteins
and the negative control. m = size marker, L = lysate. b
Silver stained SDS-PAGE gel of desalted peak fractions used
in subsequent nuclease assays. WT = wild type DmWRNexo,
TL = DmWRNexoD229V mutation (‘‘TILLING’’), DE = dou-
ble mutant DmWRNexoD162A,E164A, ‘‘M’’ = mock i.e. lysate
from E. coli transformed with the empty vector
Fig. 2 Molecular modelling of DmWRNexo to show muta-
tions generated a DmWRNexo predicted structure (see
‘‘Methods’’) is shown as a surface fill model, in green. b The
structure of human WRN exonuclease domain as reported by
Perry et al. (2006) is shown as a cyan mesh. c Merged
structures of human and Drosophila WRN exonuclease.
Individual residues are coloured as: Blue: Drosophila D162,
equivalent to D82 (human); Magenta E164 (fly) E84 (human);
Red D222 (fly) D143 (human); Yellow D229 (fly), D150
(human)
Biogerontology
123

minor band of 51kDA (asterisk in Fig. 3a) was also
sequenced and found to contain a slightly shorter
version of DmWRNexo starting from an internal
methionine eight residues downstream from the
initiator methionine.
As a negative control, a mock expression was
conducted in the same E. coli strain transformed with
the empty pIVEX2.3d vector, induced and harvested
under identical conditions to wt DmWRNexo. This
negative control bacterial lysate was applied to a His-
trap column and fractions eluted and processed as for
the recombinant his-tagged proteins; no appreciable
protein bands were detected in these ‘‘peak’’ fractions
(Fig. 3a bottom panel, lane ‘‘M’’ Fig. 3b).
DmWRNexo possesses 30?50 exonuclease
activity
To assay possible nuclease activity of the recombinant
DmWRNexo proteins, a 50 fluorescein-end labelled
oligonucleotide was annealed to complementary oli-
gonucleotides to generate duplex templates with a 50
overhang or blunt ends (see Table 1 for sequences). A
single stranded 50 fluorescein-labelled oligonucleotide
was also used to determine if the proteins exhibited
nuclease activity on single-stranded DNA. The
recombinant DmWRNexo proteins were incubated
with the various templates for 30 min then products
analysed on denaturing acrylamide gels and visualised
by excitation at 470 nm and emission at 520 nm.
From Fig. 4a, it is apparent that wt DmWRN exo
(‘‘WT’’ in Fig. 4a) possesses exonuclease activity on
both single stranded and 50 overhang duplex tem-
plates, but has no activity against a blunt-ended
template. Whilst 30–50 cleavage of the labelled
oligonucleotide was observed in duplexes with either
50 or 30 overhangs (Fig. 4a and data not shown), no
clipping of the 50 terminal fluorescein was detected.
These data, together with further analysis using other
templates (data not shown), lead us to conclude that
DmWRNexo possesses 30–50 exonuclease activity
similar to human WRN. This is the first demonstration
that the fly DmWRNexo protein is indeed an
exonuclease.
To control for the possibility that the nuclease
activity observed for wild type DmWRNexo may
result from the presence of co-purifying E. coli
Fig. 4 Nuclease assays of DmWRNexo. Fluorescently 50 end-
labelled oligonucleotides were incubated with purified
recombinant DmWRNexo and products analysed by denaturing
acrylamide gel electrophoresis. a Activity of the various forms
of DmWRNexo on various DNA substrates at 37C for 30 min
(see Table 1) for details of substrate sequences). b Impact of
temperature on DmWRNexo exonuclease activity with time (in
minutes, shown above each lane)- = buffer alone negative
control, M = mock i.e. lysate from E. coli transformed with
the empty vector, WT = wild type DmWRNexo, TL =
DmWRNexoD229V mutation (‘‘TILLING’’), DE = double
mutant DmWRNexoD162A,E164A. c Comparative nuclease
activity on the 50 overhang template at 37C of wild type
DmWRNexo (WT) versus the DmWRNexoD229V mutation
(‘‘TILLING’’) and double mutant DmWRNexoD162A,E164A
(‘‘Dead’’), as quantitated from three independent experiments
using ImageJ. Error bars are ±SEM
Biogerontology
123

nucleases, the fractions from lysate of E. coli
transformed with the empty pIVEX2.3d vector,
equivalent to the peak fractions of wt DmWRNexo
(See ‘‘M’’ in Fig. 3), were assayed for nuclease
activity. It is clear from Fig. 4a that no nuclease
activity on the assayed templates was present in these
‘‘mock’’ fractions (‘‘M’’ in Fig. 4a), allowing us to
conclude that any observed nuclease activity is
dependant upon the presence of recombinant
DmWRNexo.
Mutation decreases or ablates nuclease activity
The impact of the D229V mutation (‘‘TL’’ in Fig. 4)
on nuclease activity was analysed. Surprisingly for an
amino acid change outside the presumed active site,
DmWRNexoD229V mutant protein showed little
nuclease activity on the single stranded 50 labelled
template (left panel Fig. 4a), nor did it show any
activity on blunt duplex DNA (middle panel Fig. 4a).
Limited degradation of a duplex template with a 50
overhang was observed, but the degree of degradation
by DmWRNexoD229V after 30 min at 37C was far
less than that observed for wild type DmWRNexo
(compare WT and TL in right panel of Fig. 4a).
DmWRNexo in which the presumptive active site
residues D162 and E164 were mutated to alanine
completely lacks exonuclease activity on single
stranded or duplex templates (Fig. 4a, lanes ‘‘DE’’),
consistent with structural predictions that these
represent active site residues important for metal
ion co-ordination during catalysis (Perry et al. 2006;
Saunders et al. 2008). Thus active site mutation
completely abolished nuclease activity on the tem-
plates tested, while DmWRNexoD229V protein has
very limited 30–50 exonuclease activity.
Temperature dependence of DmWRNexo activity
Drosophila lab cultures are routinely maintained at
25C, with viability of the CG7670e04496 homozy-
gous mutant greater at 18C (data not shown). We
therefore tested the nuclease activity of DmWRNexo
at these two temperatures in addition to 37C, over a
time course of 40 min. Limited activity of the wild
type protein was detected at 18C, with increased
cleavage of the 50 overhang template at 25C and
digestion to completion at 37C after 40 min (Fig. 4b
top panel). This is consistent with an enzyme
optimum above 25C. By contrast, essentially no
cleavage of template was observed for DmWRN-
exoD229V (TL in Fig. 4b) at either 18C or 25C.
Moreover, only relatively large cleavage products
were observed on exposure of the 50 overhang
substrate to the DmWRNexoD229V mutant protein at
37C, and there was no further increase after 20 min
in either amount of cleavage product or a decrease in
size of product with time (middle panel Fig. 4b),
suggesting that this single point mutation results in an
exonuclease with very poor processivity. As antici-
pated from the data in Fig. 4a, DmWRNexo in which
two presumptive active site residues (D162, E164)
are both mutated has no observable nuclease activity
in these assays (bottom panel Fig. 4b). Simple
quantification of relative exonuclease activity of wild
type DmWRNexo and the mutant forms was calcu-
lated by taking the relative (compared to t0)
percentages of degraded compared to full-length
substrate (Fig. 4c). This method may underestimate
total activity as it ignores proportional calculation of
sequential degradation of the same strand to produce
shorter products. However, even using this relatively
crude measure, wild type DmWRNexo can robustly
turn over at least 6 9 more substrate molecules than
the DmWRNexoD229V mutant over the same time
frame at 37C (P = 0.00055, Table 2). Under these
conditions, the difference in activity between wild
type DmWRNexo and the active site mutant
DmWRNexoD162A,E164A is also very significant, as
is that between the active site mutant and DmWRN-
exoD229V (Table 2). This analysis reinforces the
results shown in Fig. 4a, b, in that wild type
DmWRNexo possesses significant and processive
exonuclease activity, whilst DmWRNexoD229V has
very little activity and the active site mutant
DmWRNexoD162A,E164A has essentially no activity
under these experimental conditions.
Discussion
Werner syndrome provides a very useful model in
which to study human ageing, but suffers from
several drawbacks. Firstly, as it is a very rare
syndrome, patient-derived material is limited in both
quantity and cell type available. Genetic heterogene-
ity between patients represents a further challenge to
analysis, and the presence of both helicase and
Biogerontology
123

exonuclease activities on the same polypeptide makes
interpretation of point mutation or RNAi experiments
complex. We are therefore developing a fly model of
Werner syndrome, since this allows a full dissection
of the role of WRN through development of the
organism, together with the ability to correlate
phenotype in vivo with biochemical activity of the
encoded protein in vitro. In invertebrates, as in
bacteria (Courcelle and Hanawalt 1999) and plants
(Hartung and Puchta 2006), we propose that the
exonuclease and helicase activities of WRN are
encoded on distinct genetic loci, enabling relatively
easy genetic dissection of their roles. To this end, we
have identified the Drosophila orthologue of the
human exonuclease activity of WRN, encoded by the
gene CG7670, and shown a hyperrecombinant phe-
notype in flies with a severe hypomorphic allele of
CG7670 in which very little mRNA is expressed due
to piggyBac insertion into the 50 UTR of the gene
(Cox et al. 2007; Saunders et al. 2008). DmWRNexo,
like WRN in higher eukaryotes, is therefore
important in preventing excessive mitotic recombi-
nation that may arise as a consequence of DNA
double strand breaks, for example at stalled replica-
tion forks (Rodriguez-Lopez et al. 2007). We are
therefore now investigating how the WRN exonucle-
ase suppresses or otherwise modulates mitotic
recombination. Here, we have utilised TILLING to
obtain a point mutation in CG7670, in an attempt to
determine what features of DmWRNexo are impor-
tant in its anti-recombinational role.
Interestingly, we find that flies heterozygous
for CG7670D229V and CG7670e04496 (mwh1
CG7670e04496/CG7670D229V) have elevated levels of
recombination that are *20-fold higher than back-
ground recombination in controls, although such
recombination frequencies are dwarfed by those
observed for the hypomorphic CG7670e04496 homo-
zygotes, and even more so by flies hemizygous for
CG7670e04496 (Fig. 1). These data strongly support
the suggestion that DmWRNexo is limiting in flies
with the piggyBac insertion in the CG7670 gene, and
that the degree of recombination is inversely related
to the amount of DmWRNexo protein present: it
probably acts stoichiometrically with DNA recombi-
nation substrates or intermediates in vivo.
Interestingly, whilst the D229V mutant retains some
ability to restrain excessive mitotic recombination,
this is partially defective.
To understand why the D229V point mutation of
DmWRNexo has an impact on DNA recombination,
we turned to molecular modelling to identify the
possible role of this amino acid in the protein. Such
modelling (Fig. 2) suggests that aspartate 229 does
not lie in proximity to the presumptive active site of
the protein, but rather towards the outside of the
molecule and is not predicted to be involved directly
in exonuclease catalysis. On spatial rotation of this
model (data not shown) it appears that aspartate 229
may lie in a cleft within the protein, and it is
conceivable that this may play a role in modulating
exonuclease activity, perhaps through mediating
association either with the DNA template or with
proteins that may bind to DmWRNexo.
To determine if the increased rates of recombina-
tion observed in flies with the CG7670D229V mutation
might arise from decreased exonuclease activity of
DmWRNexoD229V, or some other factor such as loss
of a protein interaction site, we developed a novel
fluorescence-based platform for conducting WRN
Table 2 Quantitative analysis of DmWRNexo exonuclease
activity at 37C
a
Exp 1 Exp 2 Exp 3
WT 53.42 60.92 71.82
TL 8.64 6.17 8.44
DE -1.60 -3.65 -2.63
b
t test pairs P-value
WT–DE 2.74 9 10-4
WT–TL 5.50 9 10-4
TL–DE 4.68 9 10-4
WT = wild type DmWRNexo; TL = DmWRNexoD229V
(TILLING mutation); DE = DmWRNexoD162A,E164A (active
site mutant). (a) Relative percentage of exonuclease activity;
data from three independent experiments (exp). Note that these
values are relative to the zero time point for each protein. For
Image J quantification, the total signal was calculated as the
sum of the signals from regions of the gel assigned as ‘‘uncut’’
and ‘‘degraded’’; the proportion of degraded to total is shown
above (part a) and in Fig. 4c. In nuclease-active samples (WT),
the background ‘‘shadow’’ is lower due to rapid exonuclease
activity; with little or no active nuclease (TL and DE), negative
values arise from the higher amount of ‘‘shadow’’ at zero time.
(b) Statistical analysis of these data: P values were calculated
for pairwise comparisons using a 2-tailed student t test with
v = 2
Biogerontology
123

exonuclease assays, which may be adaptable to
higher throughput assays. Importantly, we show for
the first time that DmWRNexo does have exonucle-
ase activity on both single stranded and duplex
templates bearing overhangs (but not blunt ended
duplex DNA), and moreover that it cleaves in a
30?50 direction, as reported for human WRN (Huang
et al. 1998). Furthermore, mutation of acidic amino
acids D162 and E164 to alanine, equivalent to the
catalytically critical metal–ion coordinating residues
in human WRN, ablates all exonuclease activity,
further strengthening the similarity with human
WRN. Interestingly, recombinant DmWRNexo bear-
ing the D229V mutation showed very limited
exonuclease activity on a 50 overhang substrate but
absolutely no activity was detectable against single
stranded DNA. Moreover, the small amount of
cleavage detected with this mutant enzyme did not
increase with time, and was only detectable at 37C, a
non-physiological temperature for flies, though it is
important to note that in the wild, fly larvae will often
unavoidably experience relatively high temperatures,
although not often over 37C for sustained periods. It
is possible that whilst flies are maintained at 25C in
the lab, DmWRNexo optimum is higher than this—
many mammalian enzymes have increased turnover
rates at temperatures above physiological until they
begin to denature and lose activity. In terms of the
exonuclease assay, it is also conceivable that greater
‘‘breathing’’ of duplex DNA will occur at 37C
compared with 25C, allowing access to single
stranded DNA ends for enzyme binding: however,
the results using a blunt duplex template at 37C
(Fig. 4a) do not support this assertion.
We can only speculate at this stage on why
mutation of amino acid aspartate 229, distant from
the active site, has such a marked impact on
exonuclease activity. Alteration of the charged
aspartate 229 on the protein surface to a non-polar
valine residue may impact significantly on the folding
or stability of the protein in vitro, while protein
interactions around this site in vivo may act to
stabilize the mutant form, accounting for the low
nuclease activity in vitro but the mild phenotype in
flies. Alternatively, it is possible that the DNA
substrate loops over this region of the protein prior
to entering the active site and that changes in surface
charge of the protein negatively impact on such
interactions—perhaps a duplex template interacts
more stably than a single stranded template on the
altered protein surface. The low processivity of the
DmWRNexoD229V mutant enzyme suggests that
DNA-protein association may be important: it is
highly possible that the substrate simply dissociates
from the D229V mutant form of the enzyme after
cleavage of a few nucleotides, because association is
too weak. These speculations await experimental
confirmation in the form of DNA-enzyme co-crystals,
or structural investigation by sensitive techniques
such as NMR. What must also be considered is the
severe loss of nuclease activity in vitro of the D229V
mutant enzyme, compared with a relatively mild
recombination phenotype in vivo for flies with the
CG7670e04496/CG7670D229V genotype. From the
results in Fig. 1a, b, it is clear that DmWRNexo
protein is limiting in CG7670e04496 homozygotes,
such that recombination rates (single cell clones with
mwh phenotype) approximately double when the
gene dosage halves. We assume that mutation of the
D229 residue does not impact on levels of protein
expression, so that although flies with the mutation
may have little exonuclease activity per molecule,
they will have appreciable amounts of the enzyme. It
is therefore possible that there is still sufficient
residual exonuclease activity for excessive Holliday
junction-dependent recombination to be suppressed
(see Rodriguez-Lopez et al. 2007). Additionally, it is
highly likely that DmWRNexo does not operate in
isolation within the nucleus, and that it is part of a
large multiprotein recombination and/or replication
complex, as shown for human WRN (Lebel et al.
1999). It is possible that in modulating recombina-
tion, DmWRNexo’s most important role is
recruitment of other factors, such as RecQ helicases
(e.g. DmBLM, encoded by mus309 (McVey et al.
2007)), rather than exonuclease activity per se. We
are therefore currently analyzing DNA recombination
involving Drosophila RecQ helicases. In conclusion,
using tagged recombinant DmWRNexo, we have
shown for the first time that DmWRNexo possesses
30–50 exonuclease activity, and generated an active
site mutant version of the enzyme which lacks
nuclease activity. In addition, we have correlated
increased mitotic recombination in flies bearing a
single D229V mutation in DmWRNexo with a large
reduction in nuclease activity in vitro. These data
strongly support the importance of DmWRNexo in
suppressing genomic instability in flies.
Biogerontology
123

Acknowledgments We thank Christine Borer for technical
support, and Tony Willis for protein sequencing. This work
was supported by the Biotechnology and Biological Sciences
Research Council [grant numbers BB/E000924/1, BB/
E002072/1 and BB/E016995/1].
References
Cheng WH, Muftuoglu M, Bohr VA (2007) Werner syndrome
protein: functions in the response to DNA damage and
replication stress in S-phase. Exp Gerontol 42:871–878.
doi:10.1016/j.exger.2007.04.011
Courcelle J, Hanawalt PC (1999) RecQ and RecJ process
blocked replication forks prior to the resumption of rep-
lication in UV-irradiated Escherichia coli. Mol Gen Genet
262:543–551. doi:10.1007/s004380051116
Cox LS, Faragher RG (2007) From old organisms to new
molecules: integrative biology and therapeutic targets in
accelerated human ageing. Cell Mol Life Sci 64:2620–
2641. doi:10.1007/s00018-007-7123-x
Cox LS, Clancy DJ, Boubriak I, Saunders RD (2007) Modeling
Werner Syndrome in Drosophila melanogaster: hyper-
recombination in flies lacking WRN-like exonuclease.
Ann N Y Acad Sci 1119:274–288. doi:10.1196/annals.
1404.009
Davis T, Baird DM, Haughton MF, Jones CJ, Kipling D (2005)
Prevention of accelerated cell aging in Werner syndrome
using a p38 mitogen-activated protein kinase inhibitor. J
Gerontol A Biol Sci Med Sci 60:1386–1393
Dhillon KK, Sidorova J, Saintigny Y, Poot M, Gollahon K,
Rabinovitch PS, Monnat RJ Jr (2007) Functional role of
the Werner syndrome RecQ helicase in human fibroblasts.
Aging Cell 6:53–61. doi:10.1111/j.1474-9726.2006.
00260.x
Faragher RG, Kill IR, Hunter JA, Pope FM, Tannock C, Shall
S (1993) The gene responsible for Werner syndrome may
be a cell division ‘‘counting’’ gene. Proc Natl Acad Sci
USA 90:12030–12034. doi:10.1073/pnas.90.24.12030
Hartung F, Puchta H (2006) The RecQ gene family in plants. J
Plant Physiol 163(3):287–296. doi:10.1016/j.jplph.2005.
10.013
Hayflick L (1965) The limited in vitro lifetime of human
diploid cell strains. Exp Cell Res 37:614–636. doi:
10.1016/0014-4827(65)90211-9
Huang S, Li B, Gray MD, Oshima J, Mian IS, Campisi J (1998)
The premature ageing syndrome protein, WRN, is a 30?50
exonuclease. Nat Genet 20:114–116. doi:10.1038/2410
Lawrence P, Johnston P, Morata G (1986) Methods of marking
cells. In: Roberts DB (ed) Drosophila: a practical
approach. IRL, Oxford, pp 229–242
Lebel M, Spillare EA, Harris CC, Leder P (1999) The Werner
syndrome gene product co-purifies with the DNA
replication complex and interacts with PCNA and topoi-
somerase I. J Biol Chem 274(53):37795–37799. doi:
10.1074/jbc.274.53.37795
McCallum CM, Comai L, Greene EA, Henikoff S (2000)
Targeting induced local lesions IN genomes (TILLING)
for plant functional genomics. Plant Physiol 123:439–442.
doi:10.1104/pp.123.2.439
McVey M, Andersen SL, Broze Y, Sekelsky J (2007) Multiple
functions of Drosophila BLM helicase in maintenance of
genome stability. Genetics 176:1979–1992. doi:
10.1534/genetics.106.070052
Opresko PL, Laine JP, Brosh RM Jr, Seidman MM, Bohr VA
(2001) Coordinate action of the helicase and 30 to 50
exonuclease of Werner syndrome protein. J Biol Chem
276:44677–44687. doi:10.1074/jbc.M107548200
Peitsch MC, Herzyk P, Wells TN, Hubbard RE (1996) Auto-
mated modelling of the transmembrane region of G-
protein coupled receptor by Swiss-model. Receptors
Channels 4:161–164
Perry JJ, Yannone SM, Holden LG, Hitomi C, Asaithamby A,
Han S, Cooper PK, Chen DJ, Tainer JA (2006) WRN
exonuclease structure and molecular mechanism imply an
editing role in DNA end processing. Nat Struct Mol Biol
13:414–422. doi:10.1038/nsmb1088
Rodriguez-Lopez AM, Jackson DA, Iborra F, Cox LS (2002)
Asymmetry of DNA replication fork progression in
Werner’s syndrome. Aging Cell 1:30–39. doi:10.1046/j.
1474-9728.2002.00002.x
Rodriguez-Lopez AM, Whitby MC, Borer CM, Bachler MA,
Cox LS (2007) Correction of proliferation and drug sen-
sitivity defects in the progeroid Werner’s syndrome by
Holliday junction resolution. Rejuvenation Res 10:27–40.
doi:10.1089/rej.2006.0503
Salk D, Au K, Hoehn H, Stenchever MR, Martin GM (1981)
Evidence of clonal attenuation, clonal succession, and
clonal expansion in mass cultures of aging Werner’s
syndrome skin fibroblasts. Cytogenet Cell Genet 30:108–
117. doi:10.1159/000131597
Saunders RD, Boubriak I, Clancy DJ, Cox LS (2008) Identi-
fication and characterization of a Drosophila ortholog of
WRN exonuclease that is required to maintain genome
integrity. Aging Cell 7:418–425. doi:10.1111/j.1474-
9726.2008.00388.x
Schwede T, Kopp J, Guex N, Peitsch MC (2003) SWISS-
MODEL: An automated protein homology-modeling
server. Nucleic Acids Res 31:3381–3385. doi:10.1093/
nar/gkg520
Sidorova JM, Li N, Folch A, Monnat RJ Jr (2008) The RecQ
helicase WRN is required for normal replication fork
progression after DNA damage or replication fork arrest.
Cell Cycle 7:796–807
Till BJ, Colbert T, Tompa R, Enns LC, Codomo CA, Johnson
JE, Reynolds SH, Henikoff JG, Greene EA, Steine MN,
Comai L, Henikoff S (2003) High-throughput TILLING
for functional genomics. Methods Mol Biol 236:205–220
Yu CE, Oshima J, Fu YH, Wijsman EM, Hisama F, Alisch R,
Matthews S, Nakura J, Miki T, Ouais S, Martin GM,
Mulligan J, Schellenberg GD (1996) Positional cloning of
the Werner’s syndrome gene. Science 272:258–262. doi:
10.1126/science.272.5259.258
Yu CE, Oshima J, Wijsman EM, Nakura J, Miki T, Piussan C,
Matthews S, Fu YH, Mulligan J, Martin GM, Schellen-
berg GD (1997) Mutations in the consensus helicase
domains of the Werner syndrome gene. Werner’s syn-
drome collaborative group. Am J Hum Genet 60:330–341
Biogerontology
123