Human Reproduction Vol.17, No.4 pp. 990–998, 2002
The spectrum of DNA damage in human sperm assessed by
single cell gel electrophoresis (Comet assay) and its
relationship to fertilization and embryo development
I.D.Morris1,3, S.Ilott2, L.Dixon1 and D.R.Brison2
1School
of Biological Sciences, University of Manchester, Manchester and 2Department of Reproductive Medicine,
St Mary’s Hospital, Manchester, UK
3To
whom correspondence should be addressed at: School of Biological Sciences, G38 Stopford Building, University of Manchester,
Oxford Rd, Manchester M13 9PT, UK. E-mail: [email protected]
BACKGROUND: The integrity of sperm DNA is important for the success of natural or assisted fertilization, as
well as normal development of the embryo, fetus and child. ICSI, by bypassing sperm selection mechanisms,
increases the risk of transmitting damaged DNA and the significance of this requires investigation. METHODS:
DNA damage in sperm from an unselected group of 60 men undergoing IVF treatment was measured by single
cell gel electrophoresis (Comet assay) and correlated with semen and treatment cycle parameters. RESULTS: Wide
spectra of sperm DNA damage were found both within and between men but no specific subgroups were identified.
Semen and treatment cycle parameters were not different in men grouped according to high or low sperm DNA
damage. However, regression analysis showed that DNA damage was positively associated with age (29–44 years),
abnormal sperm and motility and negatively associated with sperm concentration. In ICSI cycles DNA damage was
positively associated with impairment of post-fertilization embryo cleavage. CONCLUSIONS: This study contributes
to the evidence of DNA damage within sperm. High loads of DNA damage measured by the Comet assay were
predictive of failure of embryo development after ICSI. As it is likely that sperm with DNA damage contributed
to successful fertilization and in-vitro development, potential adverse effects remain to be clarified.
Key words: Comet analysis/DNA damage/fertility/human/sperm
Introduction
Male factor infertility remains a significant problem contributing to ~50% of the cases attending infertility clinics (Lamb
and Lipshultz, 2000; Oehninger, 2000). Clinical evaluation of
the contribution of the man towards the infertility of the couple
is usually confined to measures of total sperm count and
concentration, abnormal forms, motility and seminal factors
such as pH and antisperm antibodies. The predictive value of
these measurements is limited although some progress has
been made in recent years by the introduction of standardized
techniques, automation of motility estimations and the
organization of quality assurance schemes for semen analysis
(Van Voorhis and Sparks, 1999; Oehninger, 2000; Sakkas and
Tomlinson, 2000). Although these analyses may describe some
aspects of the function of the testis and sperm, they do not
address the integrity of the male genome contained in the head
of the sperm. Abnormalities in the male genome are, however,
a clear potential reason for post-fertilization failure (Lopes
et al., 1998a; Sakkas et al., 2000; Aitken and Krausz, 2001),
and, in rodent studies, embryo toxicity has been associated
with chemical- and radiation-induced sperm DNA damage
(Hales and Robaire, 1997). The development of ICSI to
990
introduce the male genome directly into the oocyte has been
a major advance in the success of assisted reproduction. ICSI
overcomes many of the factors related to the production of
adequate numbers of sperm as well as impaired fertilization
mechanisms. However, since sperm are selected for ICSI based
only on motility and gross morphology, concern has been
expressed that such methods bypass natural selection and could
inadvertently introduce a defective paternal genome (Campbell
and Irvine, 2000; Hargreave, 2000; Mortimer, 2000; Sakkas
et al., 2000). Some progress has been made in addressing this
issue and epidemiological analysis of the children arising from
ICSI has produced few data of major concern although the
risk remains controversial (Palermo et al., 2000; Tarlatzis and
Bili, 2000; Ericson and Kallen, 2001; Sutcliffe et al., 2001).
As the numbers of children conceived by ICSI is small and
they are still young it is possible that deleterious genetic effects
which are either subtle, low in frequency or are not expressed
until adulthood remain to be discovered.
To assist in the risk assessment of ICSI, it would be
appropriate to develop methods to measure DNA damage in
the sperm and to correlate this with biological outcomes. DNA
abnormalities in sperm are well documented. Cytogenetic
© European Society of Human Reproduction and Embryology
DNA damage in sperm
analysis of sperm chromosomes has demonstrated sperm aneuploidy, which, although low in frequency, is associated with
infertility and adverse pregnancy outcome (Egozcue et al.,
2000; Shi and Martin, 2000). Genetic information in the sperm
genome may be mutated or deleted altogether which may be
the cause of some cases of male infertility (Hargreave, 2000;
Foresta et al., 2001). However, it has also become clear that
other subtle genetic changes may be occurring. The nature of
these is not well documented but could give rise to a spectrum
of responses including: failure of fertilization, failure of
preimplantation embryo development, early pregnancy loss or
fetal abnormalities (Hales and Robaire, 1997; Sakkas et al.,
2000; Shen and Ong, 2000).
Several techniques are available to examine the integrity of
sperm DNA. The sperm chromatin structure assay (SCSA)
measures the susceptibility of the DNA to acid denaturization
(Evenson et al., 1999). Abnormal chromatin structure is
measured by flow cytometry that records the ratio of denatured
to native DNA. High ratios correlate with sperm concentration
and sperm head abnormalities. Additionally it has been
shown that if the percentage of cells with abnormal ratios
exceeds 30–40% then fertility is unlikely (Evenson et al.,
1999; Larson et al., 2000; Spano et al., 2000). DNA
strand breaks in sperm have also been directly detected by a
variety of techniques. The terminal deoxynucleotidyl transferase-mediated dUDP nick-end labelling (TUNEL)
assay identifies double and single DNA strand breaks by
enyzmatically labelling the free 3⬘ OH end of the DNA with
a fluorescent substrate. TUNEL-positive cells are identified
either microscopically or by fluorescence-activated cell sorting
(FACS). In the majority of samples only ~5% of sperm have
TUNEL-detectable DNA damage (Manicardi et al., 1995;
Aravindan et al., 1997; Sun et al., 1997; Donnelly et al., 2000;
Irvine et al., 2000; Oosterhuis et al., 2000; Ramos and Wetzels,
2001). Sperm with DNA strand breaks may persist into the
ejaculate because of a failure in the mechanism of apoptosis,
which normally eliminates them during spermatogenesis
(Sakkas et al., 1999) or alternatively may arise in the reproductive tract as a result of oxidative damage (Aitken and
Krausz, 2001). Fertilization after ICSI or IVF is reduced if
sperm are retrieved from ejaculates containing high numbers
of TUNEL-positive cells. Additionally, it has been shown that
paternal smoking increases sperm DNA damage measured by
the TUNEL technique and this has been suggested to account
for the increase in childhood cancer (Sun et al., 1997; Potts
et al., 1999).
The Comet assay is extensively used in somatic cells to
measure genotoxic damage, especially single and double strand
breaks, and was originally applied to sperm by Singh (Singh
et al., 1989). This study demonstrated that sperm DNA was
extremely labile in the presence of alkali and it was difficult
to distinguish the level of damage between individual sperm.
However, the Comet assay has been modified and used in
other laboratories to investigate DNA damage in sperm and
its relationship to infertility. Irvine and colleagues showed that
men attending infertility clinics had a higher level of DNA
damage in their sperm, which was also negatively related to
semen concentration (Irvine et al., 2000). Similar DNA changes
could be generated by in-vitro treatments producing oxidative
damage (Aitken et al., 1998; Twigg et al., 1998a; Donnelly
et al., 1999; Shen and Ong, 2000; Ramos and Wetzels, 2001).
However, another study did not detect a difference in DNA
damage between infertile men and normal controls (Hughes
et al., 1996), although subsequent reports suggest that DNA
damage measured by the Comet assay is selectively increased
in the sperm from infertile men after cryopreservation or invitro X-ray radiation (Hughes et al., 1996; Donnelly et al.,
2001). Results from the Comet assay are also correlated with
DNA damage measured by the TUNEL and SCSA methods
(Aravindan et al., 1997; Donnelly et al., 2000). Close inspection
of these publications reveals significant differences in the
protocols used to treat the cells prior to electrophoresis in an
alkaline buffer. These methodological differences have arisen
from the difficulties encountered in releasing the DNA from
the sperm head due to the unique DNA compaction (Ward
and Coffey, 1991). Combined with the alkaline lability of
sperm DNA, the difficulties encountered in releasing the DNA
for electrophoresis have hindered the development of the
Comet assays for comparative studies of male reproduction.
Our laboratory has recently developed a neutral Comet assay
to provide quantitative measures of DNA damage in human
and murine sperm (Haines et al., 1998). This method will
produce a linear increase in DNA damage of sperm irradiated
in vitro at doses that would be expected to produce DNA
strand breakage. We have demonstrated that, in the mouse,
spermatogonial radiation with external X-rays or internal
isotopic contamination results in the appearance of substantial
damage in DNA of the sperm (Haines et al., 2001). These invivo treatments are known to be genotoxic and adversely
effect fertility and reproduction. Additionally we have also
shown that during chemotherapy using fludarabine of a patient
with chronic lymphocytic leukaemia there was a substantial
increase in the sperm DNA damage when measured by the
Comet assay (Chatterjee et al., 2000). DNA damage may be
a good biomarker to relate to fertility problems; moreover, as
the Comet assay is technically straightforward and inexpensive
it may be suitable for routine measurement of DNA damage
once validated by independent laboratories. We have therefore
examined DNA damage in men attending for IVF/ICSI treatment using the Comet assay and correlated the DNA damage
profiles in their semen with treatment cycle outcome.
Materials and methods
Patient selection
Sixty couples requesting IVF treatment at St Mary’s Hospital,
Manchester were recruited into this study after providing written
informed consent for sperm research and with local ethics committee
approval. All women were aged ⬍40 years, men ⬍55 years, and the
couples were childless. Forty of these couples were selected for ICSI
according to guidelines developed at St Mary’s Hospital, based
on semen analysis parameters (World Health Organization, 1999),
fertilization performance in previous cycles, and previous history of
spontaneous conceptions. The other 20 couples received conventional IVF.
991
I.D.Morris et al.
IVF/ICSI treatment cycles
Ovarian stimulation was achieved using a conventional long protocol
down-regulation involving pituitary desensitization with buserelin.
Exogenous FSH was administered by a step-down protocol with an
initial starting dose between 75 and 450 IU and adjusted throughout
the cycle following monitoring of serum estradiol levels. HCG was
administered when three or more follicles reached 艌17 mm, and
oocytes were recovered 36 h later by ultrasound-guided retrieval.
Semen samples were produced by masturbation and analysed for
sperm concentration (⫻106/ml), percentage of sperm which were
progressively motile, and percentage abnormal forms, then prepared
by density gradient centrifugation using standard protocols. Briefly,
sperm were washed free of seminal plasma using sperm culture
medium (Medi-Cult UK Ltd, Redhill, Surrey, UK) layered upon a
40: 80% PureSperm (Hunter Scientific, Saffron Walden, Essex, UK)
gradient, centrifuged at 600 g for 20 min and resuspended in the
sperm culture medium (Horne et al., 1997). For IVF treatment,
oocytes were inseminated 4–6 h post recovery with ~1⫻105 sperm/ml.
For ICSI treatment, single motile sperm of normal morphology were microinjected into each oocyte using standard protocols
(Van Steirteghem et al., 1993). An aliquot of each sample used for
IVF or ICSI was immediately snap-frozen at –20°C until analysed
for DNA damage by the Comet assay. Oocytes were cultured overnight
at 37°C in 200 µl of Universal IVF medium (Medi-Cult UK).
Fertilization and embryo development
Oocytes were assessed for the presence of two pronuclei 16–18 h
after insemination, indicative of normal fertilization (day 1). Up to
six normally fertilized zygotes were maintained in culture to cleave
to the 2- or 4-cell stage (day 2), whilst excess zygotes were
immediately cryopreserved. Two embryos, or occasionally one or
three, were replaced in the uterus on day 2 following selection based
on developmental stage and morphological criteria (Steer et al.,
1992). Embryos received a stage score, with one point awarded for
each intact blastomere, and a morphological grade, with 4 points
awarded for even blastomeres with no fragmentation, 3 points for
embryos with up to 10% fragmentation and so on. This scoring
system is thought to reflect anucleate fragmentation resulting from
cytokinesis and/or loss of blastomeres by apoptosis. The percentage
of embryos which failed to cleave (remain at 1-cell) was also scored.
Cryopreserved embryos were thawed and replaced in subsequent
cycles at patient request. Implantation of fresh or frozen embryos
was assessed by measuring serum β-HCG levels 14 days after
replacement, clinical pregnancy was confirmed by presence of a fetal
heart on scan at 6 weeks, and live birth data were obtained by patient
follow-up.
Comet DNA damage assay
Single cell gel electrophoresis of sperm DNA (Comet assay) was performed as previously described (Haines et al., 1998). Briefly, sperm
cells were thawed rapidly at room temperature, cast into miniature
agarose gels on microscope slides and lysed in situ to remove DNA
associated proteins and allow the compacted DNA in the sperm to
relax. Lysis buffer (Tris 10 mmol/l, 0.5 mol/l EDTA and 2.5 mol/l
NaCl, pH 10) contained 1% Triton X-100, 40 mmol/l dithiothreitol
and proteinase K, 100 µg/ml). Microgels were then electrophoresed
(20 min at 25V/0.01A) in neutral buffer (Tris 10 mmol/l containing
0.08 mol/l boric acid and 0.5 mol/l EDTA, pH 8.2), during which the
damaged DNA migrated from the nucleus towards the anode. DNA
was visualized by staining of the slides with the fluorescent DNA
binding dye SYBR Green I (Molecular Probes, Oregon, USA) and
sperm identified by size and the presence of a tail. Comet measurements
performed were tail length, tail moment and percentage tail DNA using
992
Table I. Summary of Comet measurements for the cohort of men (n ⫽ 60)
investigated in this study
Mean
SD
Median
Range
Tail moment
Tail length (µm)
% tail DNA
20.09
10.08
19.73
3.4–49.03
85.49
20.85
83.87
39.4–144.8
41.49
16.30
43.13
12.07–69.04
a Nikon Optiphot II epifluorescence microscope and Comet Assay II
software (Perceptive Instruments, Haverhill, UK); 100–120 cells were
analysed per semen sample (two duplicate sample slides, 50–100
randomly selected cells scored per slide, up to 200 sperm in total).
Measurements between the two slides were highly reproducible. For
example, when the mean moments were compared for 30 randomly
selected samples with moments ranging from 0.7 to 30, the values
between the two slides were highly correlated (P ⬍ 0.0001, r2 ⫽ 0.975).
Since intra-sample variation in this assay is so low as to be negligible,
data are presented throughout as the mean of the sperm analysed on the
two slides.
Statistical analysis
Data were analysed using SPSS software (SPSS Inc., Chicago, IL,
USA). Comet parameters were subjected to further analysis using mean
(if normally distributed) or median (if not normal) values. Principle
component analysis was used to separate sperm into two populations,
which were compared using the non-parametric Mann–Whitney Utest. Multivariate stepwise regression analysis was used to establish
relationships between Comet parameters and semen analysis or treatment cycle parameters.
Results
DNA damage in the sperm populations collected from 60
unselected men attending for IVF treatment was assessed by
the neutral Comet assay. All sperm produced measurable
Comet parameters and the summary data are presented in
Table I.
Patient variation in DNA damage profiles
Upon inspection of the DNA damage moments and tail lengths
in the sperm populations, it became clear that there was a
wide range of damage distribution profiles. Examples of low,
medium and high DNA damage profiles are given in Figures
1–3. Figure 1 shows the DNA moment frequency distribution
in sperm with a population profile that is skewed towards zero
and the median value is low. Over 50% of sperm in this
sample record moments between 5 and 10 and only a small
percentage of sperm have large moments of ⬎30. Comet tail
lengths were essentially normally distributed although there
were a few sperm with short tail lengths. Figure 2 shows an
example of a sperm population in which both the tail moment
and length were apparently normally distributed, the medians
being close to that for all samples in this study. In contrast,
some sperm populations contained a substantial number of
sperm with high levels of DNA damage (Figure 3). In this
particular sample the levels of damage measured by the tail
moment, although high, appeared normally distributed. In other
samples, however, the moments were skewed towards the
DNA damage in sperm
Figure 1. Comet analysis of sperm DNA. An example of a sperm
population with low levels of DNA damage.
Figure 3. Comet analysis of sperm DNA. An example of a sperm
population with high levels of DNA damage.
higher values with a median ~60, which is about the maximum
value that can be achieved using this method as the DNA will
not migrate further during the electrophoresis. Tail lengths
tend to be longer than in samples with lower levels of DNA
damage; in the example shown the distribution is skewed
towards the maximum.
Figure 2. Comet analysis of sperm DNA. An example of a sperm
population with medium levels of DNA damage.
Analysis of populations with low and high DNA damage
In view of these observations, the distribution data for tail
moments in each sample were analysed by principle components analysis in order to see if the DNA damage profile
data could be categorized into one or more populations. For
each sample the moments (arbitrary units) calculated for 100
sperm were allocated to size bands of 0–10, 10–20,
20–30, 30–40, 40–50 and 艌50 and the distribution data
converted to the natural log before analysis. The principal
component graph (Figure 4) shows that there is a continuum
of sperm DNA damage in the different semen samples from
low to high values rather than discrete subpopulations.
We have separated these data into groups with either low
or high sperm DNA damage profiles (population 1 and 2;
Figure 4) for further analysis. The Comet assay parameters
(median and range) for population 1 (high DNA damage; n ⫽
31) were: tail moment 33.7 (5.5–39), tail length 113 (74–118),
whereas those for population 2 (low DNA damage, n ⫽ 22)
were tail moment 17.45 (4–49) and tail length 84 (39–145).
There was a significant difference between populations 1 and
2 in tail moment (P ⬍ 0.001) but not in tail length (P ⬍ 0.08,
Mann–Whitney U-test). A comparison of the semen analysis
993
I.D.Morris et al.
Table III. Multivariate stepwise linear regression analysis of Comet tail
length and moments versus semen and treatment cycle measurements
Figure 4. Principal components analysis of the DNA damage
(Comet tail moment) in the sperml populations from 60 men.
Table II. Descriptive statistics for the male population in this study
(n ⫽ 60)
Age (years)
Abstinence (days)
Semen volume (ml)
Sperm concentration
(⫻106)
Motile sperm (%)
Abnormal sperm (%)
Fertilized oocytes (%)
Cleaved zygotes (%)
Embryo cell number
Embryo grade
Pregnancy (βHCG)
Live births
Mean
Median
Range
36.1
4.9
3.3
56.3
35
4
3
48
29–47
0–28
0.5–8
0.9–200
49.1
66.4
56.3
74.9
3.04
3.11
15/60 ⫽ 25%
12/60 ⫽ 20%
52.5
65
60
83
3.0
3.0
5–90
40–90
0–100
0–100
2.0–4.0
2.0–4.0
and treatment cycle parameters (Age, abstinence, semen
volume, sperm concentration, sperm motility, percentage
abnormal forms, percentage fertilized oocytes, percentage
cleaved embryos, embryo stage, grade and clinical pregnancies)
for samples from populations 1 and 2 was also made but none
were significantly different. Interestingly, however, percentage
fertilization, percent cleavage, embryo cell number and embryo
grade all showed a tendency to be lower in the group 1
patients. There were 15 pregnancies (25% per cycle) and 12
live births (20%), evenly distributed between groups 1 and 2.
However, of the nine group 1 pregnancies, three ended in
early miscarriage, whereas all of the six pregnancies not in
group 1 yielded live births. Table II gives the summary
statistics for all patients.
Multivariate analysis
Further investigation of the relationships between the DNA
damage tail moment or tail length and clinical measurements
were made using multivariate stepwise regression analysis. In
addition to the parameters given in Table II, some data
modifications were made. Embryo cleavage was subdivided
into two groups (50–84 and 100%) and embryo development
into three groups: (1) no embryo produced (i.e. 0% cleavage),
(2) embryo grade ⬍3.5, and (3) embryo grade ⬎3.5.
The summary results from the statistical analysis for all the
samples are given in Table III. For many parameters there
was no correlation with the measurements of DNA damage.
994
Variable
Dependent
variable
Regression
coefficient
Significance
Age (years)
Motility (%)
Age (years)
Abnormal sperm (%)
Tail
Tail
Tail
Tail
1.432
0.291
0.669
0.272
0.023
0.011
0.032
0.013
length
length
moment
moment
All the parameters included in Table II were entered into the analysis;
however, only significant correlations are included here.
However, both Comet tail length and moment were significantly
correlated with age (P ⫽ 0.023 and P ⫽ 0.032 respectively)
indicating that between the ages of 29–47 years the levels of
DNA damage in sperm increased significantly. Additionally,
tail length increased with sperm motility (P ⫽ 0.011) and
tail moment increased as the proportion of abnormal forms
increased (P ⫽ 0.013). Table IV presents the data analysed
according to whether the sperm were used for either IVF or
ICSI. IVF and ICSI men were not significantly different
in age or semen characteristics such as volume or sperm
concentration. However, ICSI samples did have a lower motility
(P ⫽ 0.02) and percentage morphologically normal forms
(P ⫽ 0.04), as expected. IVF and ICSI samples did not differ
significantly in their median tail moment or length. In the ICSI
group, the associations between DNA damage and age and
motility were preserved, whereas the correlation with abnormal
forms was lost (although this was not strong in the overall
analysis). Interestingly there was a strong inverse correlation
between sperm concentration and DNA damage. Furthermore,
analysis of embryo cleavage showed that as DNA damage
measured by the tail moment increased, embryo cleavage was
impaired. Only one significant correlation was obtained for
the IVF patients—between tail length and motility (%)—which
confirms the association seen in the other analyses. However,
as the IVF group size was small it is unlikely that the analysis
possessed the power to reveal or confirm the associations
documented in Tables III and IV (ICSI).
Discussion
The results show that the Comet method we have developed
produces an estimate of the DNA damage in every sperm
sampled (100) and that in sperm from a population of men
attending for IVF treatment there is a very wide spectrum of
DNA damage both within and between men. This spectrum
of DNA damage reflects a continuum, not discrete
subpopulations, of DNA-damaged sperm from either the
individual men (individual histograms) or the patient population (principal components analysis). This result may have
arisen from the deliberate lack of selection of men undergoing
IVF/ICSI treatment who have produced a wide spectrum of
semen profiles. Others have conducted Comet analysis of DNA
damage in human sperm and have recorded the summary
statistic of the values for the individual sperm so that the
characteristics of the sperm populations have not previously
DNA damage in sperm
Table IV. Multivariate stepwise linear regression analysis of Comet tail length and moments, and semen
analysis for treatment cycles separated into either IVF or ICSI
Variable
Dependent variable
Regression
coefficient
Significance
ICSI (n ⫽ 40)
Age (years)
Semen concentration/ml (square root)
Motility (%)
Age (years)
Semen concentration/ml (square root)
Cleavage (50–84%)
Cleavage (100%)
IVF (n ⫽ 20)
Motility (%)
Tail
Tail
Tail
Tail
Tail
Tail
Tail
1.611
⫺2.647
0.305
0.666
⫺1.144
⫺7.568
⫺12.77
0.003
0.002
0.036
0.026
0.005
0.0271
0.0003
0.52
0.027
length
length
length
moment
moment
moment
moment
Tail length
All the parameters included in Table I were entered into the analysis; however, only significant correlations
are included here.
been recognized. Additionally, the method of scoring the Comet
data varies between laboratories, which could potentially affect
the resolution of the technique. Using some protocols, the
presence or absence of a Comet tail is scored, some semen
samples producing hardly any ‘Cometed’ sperm (Aravindan
et al., 1997). In other protocols, including that reported here,
all sperm produced a Comet, so allowing the images to be
analysed for % tail DNA, tail length or tail moment. The
percentage of DNA which migrates from the sperm head into
the Comet tail has been reported to be ~35% in fresh semen,
falling to 10–15% in prepared washed samples (Hughes et al.,
1997; Irvine et al., 2000; Donnelly et al., 2001). In our current
study of sperm prepared by density gradient centrifugation,
the percentage tail DNA ranged from 12 to 69%. We and
others have previously used the length of the tail to detect the
effects of in-vitro irradiation and the effects of chemotherapy
upon DNA damage in sperm (Haines et al., 1998; Singh and
Stephens, 1998; Chatterjee et al., 2000). Although others have
measured tail length, percentage DNA and moment, only
results from the percentage tail DNA measurements were
documented, presumably because this measure gave better
discrimination in their analysis (Irvine et al., 2000). Although
all three measurements are interrelated we have chosen to
present our data as tail length and tail moment, because in
this study these measures exaggerated the differences between
patients. The differences in DNA damage between ejaculates
were most impressive using the Comet tail moment (see
Figures 1–3). This may perhaps be expected as both the length
of migration of the DNA and the amounts released from the
head are independently increased by DNA damage (Olive,
1999).
It is also important to consider the type of DNA damage
the Comet technique is measuring, especially as the methods
used for analysis of sperm vary between laboratories. Comet
analysis of somatic cells electrophoresed under alkaline conditions measures single DNA strand breaks, whereas neutral
conditions reportedly measure double strand breaks. However,
many forms of structural damage can be converted to strand
breaks during cell preparation and electrophoresis (Collins
et al., 1997; Olive, 1999). Early attempts to apply the Comet
assay to sperm under alkaline electrophoresis conditions to
measure single strand DNA breaks resulted in the majority of
DNA migrating into the Comet tail (Singh et al., 1989). This
also occurs using the protocol we have used here and has been
ascribed to the abundant alkaline-sensitive sites in the sperm
DNA and reflects the fragility of the structure in vitro. More
recently, several methods have been developed which rely
upon harsh biochemical treatments, with or without proteinase
K and reducing agents. These lyse the sperm cell, allowing
the DNA to decondense by removing nuclear proteins and
their crosslinks. The differences in protocols are increased by
the electrophoresis conditions that range from pH 8.2 to 13
(Haines et al., 1998; Singh and Stephens, 1998; Hughes et al.,
1999; Irvine et al., 2000; Donnelly et al., 2001). In spite of
this, some common results have been reported, especially the
observation that radiation of sperm, which is known to induce
double and single strand breaks, will also increase the Comet
measurements (Haines et al., 1998; Singh and Stephens, 1998;
Hughes et al., 1999). It is therefore likely that although the
final measurement reflects DNA strand breaks in vitro, it is
likely that these have arisen from a variety of in-situ DNA
abnormalities ranging from strand breaks per se but also
including structural abnormalities. Thus Comet analysis of
sperm is most appropriately described generically as DNA
damage. Which of the different protocols is the most useful
for describing clinically relevant sperm DNA damage remains
to be determined.
Whilst this and other studies have established that DNA
damage is present in sperm, the biological significance of
reproduction with these sperm is not clear. It may be predicted
that high loads of DNA damage would be reflected in abnormal
fertilization and development, ultimately leading to death of
the embryo. In an attempt to address this question, we
compared the semen analysis parameters, fertilization, embryo
development and pregnancy outcome, from clinical treatment
cycles for men grouped according to whether the sperm carried
low or high DNA damage loads. Surprisingly, there were no
significant differences in any of our measurements. It may be
noteworthy that there was a trend for men in the higher DNA
damage group to show poorer results in all of the development
parameters and that early pregnancy loss occurred only in this
group. These associations are potentially interesting and need
995
I.D.Morris et al.
to be confirmed by studying a larger, more clearly defined
group of men.
However, multivariate analysis of the entire population of
samples revealed some interesting correlations. DNA damage
increased as a function of age, motility and abnormal forms.
When only the men selected for ICSI treatment were analysed,
the association with age and motility remained, but additionally semen sperm concentration and embryo cleavage were
significantly negatively associated with DNA damage. The
association of increased sperm DNA damage with abnormal
forms and decreasing sperm count confirms other studies (Sun
et al., 1997; Irvine et al., 2000) and is perhaps predictable as
inefficiency in spermatogenesis may lead to cells appearing in
the ejaculate which either have not completed development or
have escaped mechanisms to delete them within the testicular
parenchyma (Sakkas et al., 2000). There have been several
studies investigating changes in male reproduction with age
(Plas et al., 2000; Kidd et al., 2001; Paulson et al., 2001; Rolf
and Nieschlag, 2001). Whilst it is clear that endocrine activity
in ageing men is less efficient it appears that in many
studies sperm production is maintained, although there may
be increases in abnormal forms and a decline in motility. We
did not detect an age-related decline in sperm count motility
or abnormal forms, possibly because the setting of this study
resulted in a heterogeneous selection of men with a limited
age range of 29–44 years. However, the increase in DNA
damage was strongly significant. DNA fragmentation measured
by TUNEL is not correlated with age (Sun et al., 1997)
although another study has suggested that DNA chromatin
condensation is abnormal in sperm from ageing men (Haidl
et al., 1996). However, it is fairly clear that conception with
sperm from ageing men does not adversely effect fertilization
or live birth rate.
Association between sperm DNA damage and impaired
fertility has previously been reported. Sperm with high levels
of TUNEL labelling are more often found in infertile men, fail
to decondense after ICSI and fertilization is often unsuccessful
(Sakkas et al., 1996; Lopes et al., 1998b; Host et al., 2000).
The SCSA has shown that those men with high levels of
chromatin abnormalities in their sperm are likely to have poor
fertility, which may be related to impaired fetal development
and subsequent miscarriages (Evenson et al., 1999; Larson
et al., 2000; Spano et al., 2000). Although it has been shown
that there is some correlation between the DNA damage assays,
as it is not clear what structural changes these are measuring,
it may not be helpful to compare outcomes until further work
is done. However, the Comet assay has been used to compare
DNA damage in sperm from fertile and infertile men. No
differences in sperm Comet parameters between fertile and
infertile men could be found in one report (Hughes et al.,
1996), however, the study did reveal differences after in-vitro
treatment with either X-rays or hydrogen peroxide. This work
suggested that sperm DNA from the infertile men was more
susceptible to damage arising from oxidative free radical
generation (Hughes et al., 1996). This apparent increase in
DNA fragility has also been shown by freeze–thawing of
samples for fertile and infertile men (Donnelly et al., 2001).
The Comet assay has also detected significant correlation
996
between DNA damage and semen parameters in an unselected
population of men attending an infertility clinic (Irvine et al.,
2000). The latter study also highlighted significant differences
in DNA damage between infertility patients and normospermic
donors. In the combined analysis of the donors and patients
there was considerable overlap of the DNA damage profiles
of the two groups. However, there was a highly significant
increase in DNA damage as the sperm count and the proportion
of morphologically normal forms decreased. The present study
confirms the associations between DNA damage and sperm
concentration and abnormal forms.
Sperm motility is often used as a predictive measure in
semen analyses, high motility being a prerequisite of normal
sperm parameters. It was somewhat surprising therefore to
find that the higher the motility of sperm in semen, the higher
the DNA damage load carried by the sperm populations. In
contrast, others (Barroso et al., 2000; Irvine et al., 2000; Zini
et al., 2001) report that sperm samples with low motility
carried higher loads of DNA damage (TUNEL or Comet)
(Irvine et al., 2000). ROS generated in vitro decreases motility
as well as inducing sperm DNA damage (Aitken et al., 1998;
Twigg et al., 1998a; Donnelly et al., 1999; Shen and Ong,
2000; Ramos and Wetzels, 2001). DNA damage has, however,
been both negatively and positively associated with sperm
ROS production (Barroso et al., 2000; Irvine et al., 2000).
These studies suggest that while nuclear DNA damage may
be induced by ROS, at the same time low levels may promote
sperm motility (explaining our correlation) whereas higher
pathological/pharmacological concentrations impair motility
(de Lamirande et al., 1997), explaining the inverse correlation.
The most important observation from this study was that
sperm containing high loads of DNA damage detected by the
Comet assay gave rise to pronuclei at a normal incidence but
were associated with an increase in the percentage of embryos
that failed to develop after ICSI. This result is consistent with
experiments in which DNA damage in human sperm was
created artificially in vitro and after injection into hamster
oocytes pronuclear formation was unchanged (Twigg et al.,
1998b). DNA damage measured by the TUNEL assay is also
negatively related to both fertilization and embryo cleavage
rate after IVF, suggesting that DNA damaged sperm will
fertilize less efficiently and confirming that early embryonic
development is impaired (Sun et al., 1997). The results of our
study suggest that selection of the sperm for ICSI in terms of
DNA damage was random and that a mechanism for the
screening of sperm carrying damaged DNA operates after ICSI
to ensure that only those zygotes with a relatively intact
genome go on to develop. It would be reassuring to conclude
therefore that implantation and pregnancy outcome would not
be so adversely affected by using sperm samples carrying high
loads of DNA damage. However, the size of this study does
not allow us to make this conclusion, and moreover, it remains
possible that low, sublethal levels of sperm DNA damage are
transmitted through to embryo development. These may be
insufficient to trigger a gross response such as cell cycle arrest
or apoptosis prior to implantation, or early pregnancy failure,
but may nonetheless be expressed in fetal or post-natal development (Hales and Robaire, 1997; Sakkas et al., 2000). It would
DNA damage in sperm
be useful if screening semen samples for sperm DNA damage
could contribute towards the selection of patients for ICSI.
However, we cannot recommend this measurement at this time
because significant numbers of sperm in a sample carrying
high DNA damage loads may be genetically normal. Further
research will be necessary to see if techniques can be devised
to identify and select sperm with undamaged DNA for ICSI,
or to remove sperm with damaged DNA from sperm samples,
to enable the pregnancy outcome after ICSI to be improved.
This work may also have implications for the genetic integrity
and normal development of children conceived by IVF, especially ICSI.
Acknowledgements
We would like to thank I.Panchbhaya and C.McCann for excellent
technical support, Dr R.Swindell for statistical help and advice and
G.Horne and H.Hunter for assistance in collecting semen samples.
References
Aitken, R.J. and Krausz, C. (2001) Oxidative stress, DNA damage and the Y
chromosome. Reproduction, 122, 497–506.
Aitken, R.J., Gordon, E., Harkiss, D., Twigg, J.P., Milne, P., Jennings, Z. and
Irvine, D.S. (1998) Relative impact of oxidative stress on the functional
competence and genomic integrity of human sperm. Biol. Reprod., 59,
1037–1046.
Aravindan, G.R., Bjordahl, J., Jost, L. and Evenson, D.P. (1997) Susceptibility
of human sperm to in situ DNA denaturation is strongly correlated with
DNA strand breaks identified by single cell electrophoresis. Exp. Cell Res.,
236, 231–237.
Barroso, G., Morshedi, M. and Oehninger, S. (2000) Analysis of DNA
fragmentation, plasma membrane translocation of phosphatidylserine and
oxidative stress in human sperm. Hum. Reprod., 15, 1338–1344.
Campbell, A.J. and Irvine, D.S. (2000) Male infertility and intracytoplasmic
sperm injection (ICSI). Br. Med. Bull., 56, 616–629.
Chatterjee, R., Haines, G., Perera, D.M.D., Goldstone, A. and Morris, I.D.
(2000) Testicular and sperm DNA damage after treatment with fludarabine
for chronic lymphocytic leukaemia. Hum. Reprod., 15, 762–766.
Collins, A.R., Dobson, V.L., Dusinska, M., Kennedy, G. and Stetina, R. (1997)
The comet assay: what can it really tell us? Mutat. Res., 375, 183–193.
de Lamirande, E., Jiang, H., Zini, A., Kodama, H. and Gagnon, C. (1997)
Reactive oxygen species and sperm physiology. Rev. Reprod., 2, 48–54.
Donnelly, E.T., McClure, N. and Lewis, S.E. (1999) The effect of ascorbate
and alpha-tocopherol supplementation in vitro on DNA integrity and
hydrogen peroxide-induced DNA damage in human sperm. Mutagenesis,
14, 505–512.
Donnelly, E.T., O’Connell, M., McClure, N. and Lewis, S.E. (2000) Differences
in nuclear DNA fragmentation and mitochondrial integrity of semen and
prepared human sperm. Hum. Reprod., 15, 1552–1561.
Donnelly, E.T., Steele, E.K., McClure, N. and Lewis, S.E. (2001) Assessment
of DNA integrity and morphology of ejaculated sperm from fertile and
infertile men before and after cryopreservation. Hum. Reprod., 16, 1191–
1199.
Egozcue, S., Blanco, J., Vendrell, J.M., Garcia, F., Veiga, A., Aran, B., Barri,
P.N., Vidal, F. and Egozcue, J. (2000) Human male infertility: chromosome
anomalies, meiotic disorders, abnormal sperm and recurrent abortion. Hum.
Reprod. Update, 6, 93–105.
Ericson, A. and Kallen, B. (2001) Congenital malformations in infants born
after IVF: a population-based study. Hum. Reprod., 16, 504–509.
Evenson, D.P., Jost, L.K., Marshall, D., Zinaman, M.J., Clegg, E., Purvis, K.,
De Angelis, P. and Claussen, O.P. (1999) Utility of the sperm chromatin
structure assay as a diagnostic and prognostic tool in the human fertility
clinic. Hum. Reprod., 14, 1039–1049.
Foresta, C., Moro, E. and Ferlin, A. (2001) Y chromosome microdeletions
and alterations of spermatogenesis. Endocr. Rev., 22, 226–239.
Haidl, G., Jung, A. and Schill, W.B. (1996) Ageing and sperm function. Hum.
Reprod., 11, 558–560.
Haines, G., Marples, B., Daniel, P. and Morris, I. (1998) DNA damage in
human and mouse sperm after in vitro-irradiation assessed by the comet
assay. Adv. Exp. Med. Biol., 444, 79–93.
Haines, G.A., Hendry, J.H., Daniel, C.P. and Morris, I.D. (2001) Increased
levels of Comet-detected sperml DNA damage following in vivo isotopicor X-irradiation of spermatogonia. Mutat. Res., 495, 21–32.
Hales, B.F. and Robaire, B. (1997) Paternally mediated effects on development.
Handbook of Developmental Toxicity. CRC Press, New York, pp. 91–107.
Hargreave, T. (2000) Genetically determined male infertility and assisted
reproduction techniques. J. Endocrinol. Invest., 23, 697–710.
Horne, G., Critchlow, J.D., Newman, M.C., Edozien, L., Matson, P.L. and
Lieberman, B.A. (1997) A prospective evaluation of cryopreservation
strategies in a two-embryo transfer programme. Hum. Reprod., 12, 542–547.
Host, E., Lindenberg, S. and Smidt-Jensen, S. (2000) DNA strand breaks in
human sperm: correlation with fertilization in vitro in oligozoospermic men
and in men with unexplained infertility. Acta Obstet. Gynecol. Scand., 79,
189–193.
Hughes, C.M., Lewis, S.E., McKelvey-Martin, V.J. and Thompson, W. (1996)
A comparison of baseline and induced DNA damage in human sperm from
fertile and infertile men, using a modified comet assay. Mol. Hum. Reprod.,
2, 613–619.
Hughes, C.M., Lewis, S.E., McKelvey-Martin, V.J. and Thompson, W. (1997)
Reproducibility of human sperm DNA measurements using the alkaline
single cell gel electrophoresis assay. Mutat. Res., 374, 261–268.
Hughes, C.M., McKelvey-Martin, V.J. and Lewis, S.E.M. (1999) Human
sperm DNA integrity assessed by the Comet and ELISA assays. Mutagenesis,
14, 71–75.
Irvine, D.S., Twigg, J.P., Gordon, E.L., Fulton, N., Milne, P.A. and Aitken,
R.J. (2000) DNA integrity in human sperm: relationships with semen
quality. J. Androl., 21, 33–44.
Kidd, S.A., Eskenazi, B. and Wyrobek, A.J. (2001) Effects of male age on
semen quality and fertility: a review of the literature. Fertil. Steril., 75,
237–248.
Lamb, D.J. and Lipshultz, L.I. (2000) Male infertility: recent advances and a
look towards the future. Curr. Opin. Urol., 10, 359–362.
Larson, K.L., DeJonge, C.J., Barnes, A.M., Jost, L.K. and Evenson, D.P.
(2000) Sperm chromatin structure assay parameters as predictors of failed
pregnancy following assisted reproductive techniques. Hum. Reprod., 15,
1717–1722.
Lopes, S., Jurisicova, A. and Casper, R.F. (1998a) Gamete-specific DNA
fragmentation in unfertilized human oocytes after intracytoplasmic sperm
injection. Hum. Reprod., 13, 703–708.
Lopes, S., Sun, J.G., Jurisicova, A., Meriano, J. and Casper, R.F. (1998b)
Sperm deoxyribonucleic acid fragmentation is increased in poor-quality
semen samples and correlates with failed fertilization in intracytoplasmic
sperm injection. Fertil. Steril., 69, 528–532.
Manicardi, G.C., Bianchi, P.G., Pantano, S., Azzoni, P., Bizzaro, D., Bianchi,
U. and Sakkas, D. (1995) Presence of endogenous nicks in DNA of ejaculated
human sperm and its relationship to chromomycin A3 accessibility. Biol.
Reprod., 52, 864–867.
Mortimer, D. (2000) The future of male infertility management and assisted
reproduction technology. Hum. Reprod., 15, 98–110.
Oehninger, S. (2000) Clinical and laboratory management of male infertility:
an opinion on its current status. J. Androl., 21, 814–821.
Olive, P.L. (1999) DNA damage and repair in individual cells: applications
of the comet assay in radiobiology. Int. J. Radiat. Biol., 75, 395–405.
Oosterhuis, G.J., Mulder, A.B., Kalsbeek-Batenburg, E., Lambalk, C.B.,
Schoemaker, J. and Vermes, I. (2000) Measuring apoptosis in human sperm:
a biological assay for semen quality? Fertil. Steril., 74, 245–250.
Palermo, G.D., Neri, Q.V., Hariprashad, J.J., Davis, O.K., Veeck, L.L. and
Rosenwaks, Z. (2000) ICSI and its outcome. Semin. Reprod. Med., 18,
161–169.
Paulson, R.J., Milligan, R.C. and Sokol, R.Z. (2001) The lack of influence of
age on male fertility. Am. J. Obstet. Gynecol., 184, 818–822.
Plas, E., Berger, P., Hermann, M. and Pfluger, H. (2000) Effects of aging on
male fertility? Exp. Gerontol., 35, 543–551.
Potts, R.J., Newbury, C.J., Smith, G., Notarianni, L.J. and Jefferies, T.M.
(1999) Sperm chromatin damage associated with male smoking. Mutat.
Res., 423, 103–111.
Ramos, L. and Wetzels, A.M. (2001) Low rates of DNA fragmentation in
selected motile human sperm assessed by the TUNEL assay. Hum. Reprod.,
16, 1703–1707.
Rolf, C. and Nieschlag, E. (2001) Reproductive functions, fertility and genetic
risks of ageing men. Exp. Clin. Endocrinol. Diabet., 109, 68–74.
997
I.D.Morris et al.
Sakkas, A.D., Manicardi, G., Bizarro, D. and Bianchi, P.G. (2000) Possible
consequences of performing ICSI with sperm possessing nuclear DNA
damage. Hum. Fertil., 26, 3.
Sakkas, D. and Tomlinson, M. (2000) Assessment of sperm competence.
Semin. Med., 18, 133–139.
Sakkas, D., Urner, F., Bianchi, P.G., Bizarro, D., Wagner, I., Jaquenoud, N.,
Manicardi, G. and Campana, A. (1996) Sperm chromatin anomalies can
influence decondensation after ICSI. Hum. Reprod., 11, 837–843.
Sakkas, D., Mariethoz, E., Manicardi, G., Bizzaro, D., Bianchi, P.G. and
Bianchi, U. (1999) Origin of DNA damage in ejaculated human sperm.
Rev. Reprod., 4, 31–37.
Shen, H.M. and Ong, C.N. (2000) Detection of oxidative DNA damage in
human sperm and its association with sperm function and male infertility.
Free Rad. Biol. Med., 28, 529–536.
Shi, Q. and Martin, R.H. (2000) Aneuploidy in human sperm: a review of the
frequency and distribution of aneuploidy, effects of donor age and lifestyle
factors. Cytogenet. Cell Genet., 90, 219–226.
Singh, N.P. and Stephens, R.E. (1998) X-ray-induced DNA double-strand
breaks in human sperm. Mutagenesis, 13, 75–79.
Singh, N.P., Danner, D.B., Tice, R.R., McCoy, M.T., Collins, G.D. and
Schneider, E.L. (1989) Abundant alkali-sensitive sites in DNA of human
and mouse sperm. Exp. Cell Res., 184, 461–470.
Spano, M., Bonde, J.P., Hjollund, H.I., Kolstad, H.A., Cordelli, E. and Leter,
G. (2000) Sperm chromatin damage impairs human fertility. The Danish
First Pregnancy Planner Study Team. Fertil. Steril., 73, 43–50.
Steer, C.V., Mills, C.L., Tan, S.L., Campbell, S. and Edwards, R.G. (1992)
The cumulative embryo score: a predictive embryo scoring technique to
select the optimal number of embryos to transfer in an in-vitro fertilization
and embryo transfer programme. Hum. Reprod., 7, 117–119.
Sun, J.G., Jurisicova, A. and Casper, R.F. (1997) Detection of deoxyribonucleic
acid fragmentation in human sperm: correlation with fertilization in vitro.
Biol. Reprod., 56, 602–607.
998
Sutcliffe, A.G., Taylor, B., Saunders, K., Thornton, S., Lieberman, B.A. and
Grudzinskas, J.G. (2001) Outcome in the second year of life after in-vitro
fertilisation by intracytoplasmic sperm injection: a UK case–control study.
Lancet, 357, 2080–2084.
Tarlatzis, B.C. and Bili, H. (2000) Intracytoplasmic sperm injection—survey
of world results. Ann. N.Y. Acad. Sci., 900, 336–344.
Twigg, J., Fulton, N., Gomez, E., Irvine, D.S. and Aitken, R.J. (1998a)
Analysis of the impact of intracellular reactive oxygen species generation
on the structural and functional integrity of human sperm: lipid peroxidation,
DNA fragmentation and effectiveness of antioxidants. Hum. Reprod., 13,
1429–1436.
Twigg, J.P., Irvine, D.S. and Aitken, R.J. (1998b) Oxidative damage to DNA
in human sperm does not preclude pronucleus formation at intracytoplasmic
sperm injection. Hum. Reprod., 13, 1864–1871.
Van Steirteghem, A.C., Liu, J., Joris, H., Nagy, Z., Janssenswillen, C.,
Tournaye, H., Derde, M.P., Van Assche, E. and Devroey, P. (1993)
Higher success rate by intracytoplasmic sperm injection than by subzonal
insemination. Report of a second series of 300 consecutive treatment cycles.
Hum. Reprod., 8, 1055–1060.
Van Voorhis, B.J. and Sparks, A.E. (1999) Semen analysis: what tests are
clinically useful? Clin. Obstet. Gynecol., 42, 957–971.
Ward, W.S. and Coffey, D.S. (1991) DNA packaging and organization in
mammalian sperm: comparison with somatic cells. Biol. Reprod., 44,
569–574.
World Health Organization (1999) WHO Laboratory Manual for the
examination of human semen and cervical mucus interaction. 4th edn,
Cambridge University Press, Cambridge.
Zini, A., Bielecki, R., Phang, D. and Zenzes, M.T. (2001) Correlations
between two markers of sperm DNA integrity, DNA denaturation and DNA
fragmentation, in fertile and infertile men. Fertil. Steril., 75, 674–677.
Submitted on August 17, 2001; resubmitted on November 22, 2001; accepted on
November 29, 2001