Journal of Andrology, Vol. 21, No.

3, May/June 2000
Copyright 䉷 American Society of Andrology

Sperm Preparation Methods Andrology Lab Corner

DAVID MORTIMER totally inhibit, their fertilizing capacity (Kanwar et al,
From the Genesis Fertility Centre, Vancouver, BC, 1979). Therefore, spermatozoa for clinical procedures such
Canada. as intrauterine insemination (IUI) or IVF (and also for lab-
oratory tests of sperm fertilizing ability) must be separated
from the seminal plasma environment not only as soon as
The spermatozoa of all placental (eutherian) mammals, possible after ejaculation (allowing for the required wait
including humans, are in a protective, nonlabile state at for liquefaction) but also as efficiently as possible.
ejaculation and are incapable of fertilization even if they There are 4 basic approaches to sperm preparation: 1)
are placed in direct contact with an oocyte. Consequently, simple dilution and washing, 2) sperm migration (either
they must undergo a subsequent period of final maturation directly from liquefied semen from a suspension of
during which they acquire the capacity to interact with washed spermatozoa or from a washed sperm pellet), 3)
the oocyte–cumulus complex and achieve fertilization. ‘‘selective’’ washing procedures (using density gradients),
This process, which was discovered independently by and 4) adherence methods to eliminate debris and dead
Austin and Chang in 1951, was termed capacitation, and spermatozoa (eg, glass wool, glass beads, and Sephadex
spermatozoa in the ejaculate are prevented from under- columns). These have been extensively reviewed else-
going capacitation by one or more decapacitation factors where (Mortimer and Mortimer, 1992; Mortimer,
that are present in the seminal plasma (Yanagimachi, 1994a,b). In the late 1980s, Aitken and Clarkson (1988)
1994). Capacitation of eutherian spermatozoa is essential discovered that the centrifugal pelleting of unselected hu-
for fertilization not only in vivo but also in vitro, and man sperm populations often resulted in the generation
underlies the manipulation of spermatozoa for clinical in of free radical or reactive oxygen species (ROS) within
vitro fertilization (IVF). the sperm pellet that could adversely affect sperm func-
Not only does seminal plasma contain one or more de- tion in vitro. Moreover, I prepared a detailed review of
capacitation factors that prevent spontaneous capacitation the literature of in vitro tests of human sperm function
of spermatozoa upon ejaculation, but it also contains one and IVF that clearly demonstrated that sperm preparation
or more factors to which prolonged exposure has adverse methods that included such a centrifugation step could
effects on sperm function, including the ability to pene- impair sperm function to such an extent as to cause a
trate cervical mucus (Kremer, 1968), undergo the acro- decrease in fertilization rates and even cause fertilization
some reaction in vitro, and the fertilization process in gen- failure in more extreme cases (Mortimer, 1991). The basic
eral (Rogers et al, 1983; Mortimer and Mortimer, 1992; conclusion was that the potentially hazardous practice of
Mortimer et al, 1998). Consequently, in order for euthe- washing spermatozoa by centrifuging unselected sperm
rian spermatozoa to have the capacity to fertilize an oo- populations should be abandoned in favor of known
cyte, they must be separated from the seminal plasma, ‘‘safe’’ practices, such as direct swim-up from semen and
and hence, the separation of human spermatozoa from
density gradient centrifugation techniques. Given the as-
seminal plasma is an essential prerequisite for them to be
sociation between the term ‘‘sperm washing’’ and the
able to achieve capacitation and express their intrinsic
simple, usually repeated centrifugation and resuspension
fertilizing ability. In assisted reproductive technology
of spermatozoa in culture medium, it seems appropriate
(ART) laboratories, this need is manifested in the process
to urge that the method be reserved for that type of pro-
commonly referred to as ‘‘sperm washing,’’ in which
cess and that another term, such as sperm preparation, be
spermatozoa are somehow removed from the seminal
employed when describing processes such as swim-up
plasma and resuspended in culture medium.
migration and density gradient centrifugation.
Prolonged exposure (⬎30 minutes) to seminal plasma
The purpose of this article is to review the significant
after ejaculation can permanently diminish the fertilizing
literature on sperm preparation published since 1990 and
capacity of human spermatozoa in vitro (Rogers et al,
update the recommended approaches for human sperm
1983), and contamination of prepared sperm populations
preparation within the context of modern ART practices.
with only traces of seminal plasma can diminish, or even

Correspondence to: David Mortimer, PhD, Director of Research and

Development, Genesis Fertility Centre, 550–555 West 12th Ave, Van-
Is Centrifugation Really Harmful?
couver, BC V5Z 3X7, Canada (e-mail: [email protected]).
Received for publication January 11, 2000; accepted for publication In simple physical terms, centrifugation can be harmful
January 18, 2000. to human spermatozoa, but this does not seem to be a
357
 andr 21 316 Mp 358
File # 16em

358 Journal of Andrology · May/June 2000

problem until forces in excess of 800 ⫻ g are applied must instead be employed. Which of these approaches
(Jeulin et al, 1982). After a 1991 editorial (Mortimer, might be preferable is also discussed later.
1991), several laboratories investigated the issue of such Although the success of a sperm preparation method is
ROS-induced damage, and although some researchers ap- often assessed by its yield of motile spermatozoa, it is
proached the question scientifically using hypothesis-test- also vital that sperm preparations for clinical use should
ing experiment designs (eg, Chan and Tucker, 1992), oth- be free of any microbiological contaminants present in
ers sought to dismiss the idea by illustrating that in their semen. Other relevant considerations in choosing a meth-
particular situation—usually working primarily with non– od include its technical complexity as well as its costs in
male-factor semen samples or with poorly defined ‘‘ab- materials, apparatus, and time. Any possible exposure of
normal semen parameters’’—it was not a statistically sig- spermatozoa during preparation to deleterious influences
nificant problem (eg, Englert et al, 1992). Another, albeit that may cause iatrogenic sperm dysfunction must obvi-
unrelated, paper by Morales et al (1991), demonstrated ously be avoided at all costs.
that although Percoll gradients exhibited higher sperm re-
covery, there were no apparent differences in sperm func-
tion (zona-free hamster egg penetration, inducible acro- ‘‘Safe’’ Methods for Sperm Preparation
some reactions, and sperm–zona pellucida binding) be-
tween such spermatozoa and parallel aliquots prepared by Although the list of techniques described in this article is
wash-and-swim-up procedures. However, the study was not exhaustive, the principles used to select them can be
performed on a series of 12 semen samples from normal, applied to determine whether any apparently similar ap-
fertile men. None of these subsequent studies addressed proach is likely or not to be ‘‘safe’’ (ie, whether it avoids
the real issue by investigating patients with likely iatro- the risk of ROS generation).
genic sperm dysfunction, and all left findings such as
those of Guérin et al (1989) on a large series of IVF Direct Swim-up From Semen
patients as incontrovertible evidence of induced sperm This is the original swim-up technique that has been used
dysfunction. Our now greatly increased understanding of by investigators who study mammalian sperm physiology
the mechanisms by which morphologically abnormal since at least the 1950s. Liquefied semen is layered be-
spermatozoa with retained spermatid cytoplasm and leu- neath a culture medium (or culture medium layered over
kocytes present within the ejaculate generate free radicals the semen), and during a subsequent incubation period
in vitro (see later discussion) provides mechanistic expla- that can range from 15 to 60 minutes depending on the
nations for this problem. It will certainly not affect every application, the progressively motile spermatozoa migrate
man, especially those with (more) normal sperm quality, from the semen layer into the culture medium. The inclu-
but in the infertility clinic setting, these men are a mi- sion of this migration step is considered to be functionally
nority, and the majority of such patients must be consid- equivalent to the process by which human spermatozoa
ered to be at risk of damage to their spermatozoa during escape from the ejaculate and colonize the cervical mucus
preparation for ART. (Mortimer et al, 1982; Katz et al, 1990; Mortimer, 1995;
Therefore, the earlier conclusion remains valid: cen- Mortimer, 1997), although some differences exist because
trifugal pelleting of unselected populations of human of the different rheological characteristics between the
spermatozoa causes irreversible damage to the sperma- culture medium and midcycle cervical mucus. The selec-
tozoa that can impair—even totally destroy—their fertil- tion of spermatozoa for their motility and morphology
izing ability. Hence techniques that involve any simple during in vitro migration is highly comparable (Mortimer
sperm washing step in which semen is diluted with cul- et al, 1982), but the process might be suboptimal for clin-
ture medium and centrifuged, regardless of whether the ical applications because of differences in chromatin qual-
pellet is just washed and the spermatozoa resuspended or ity using this method.
the motile spermatozoa allowed to swim out from the The Sperm Select System (Select Medical Systems,
washed pellet, must be recognized as potentially harmful Williston, Vt) employs a high-purity preparation of 3000
and sometimes lethal to the spermatozoa being pre- kd sodium hyaluronate at a 1-mg/mL final concentration
pared—and consequently, must be abandoned for any ap- in culture medium. In a clinical IVF program, swim-up
plication that requires physiological functionality of the from semen into the hyaluronate solution gave a signifi-
prepared spermatozoa. Alternative sperm preparation cantly higher percentage of motile spermatozoa compared
methods, such as direct swim-up from liquefied semen with the traditional swim-up from a washed pellet method
(with a subsequent centrifugal washing step into fresh and, ultimately, allowed the achievement of a higher preg-
medium being safe because the ROS-generating cells nancy rate (Wikland et al, 1987). Whether these improved
have been excluded from the prepared population), some results were due specifically to the use of the hyaluronate
adherence methods, or density gradient centrifugation, or to the use of a method that did not involve the initial
Mortimer · Sperm Preparation Methods 359

centrifugal pelletting was not ascertained. More recent re- dients in the early days (eg, Gorus and Pipeleers, 1981;
search has demonstrated numerous beneficial effects of Bolton and Braude, 1984), clinical applications since the
hyaluronate upon spermatozoa (Huszar et al, 1990; Sbra- late 1980s have almost exclusively employed discontin-
cia et al, 1997), and therefore Sperm Select, especially in uous gradients (eg, Arcidiacono et al, 1983; Lessley and
view of its market focus toward office gynecologists who Garner, 1983; Dravland and Mortimer, 1985). Discontin-
perform IUI, must be considered a useful technique for uous gradients are usually prepared with 2 or 3 layers,
clinical human sperm preparation (Zimmerman et al, although methods that use up to 12 layers have been re-
1994). However, it is important to undertake additional ported for special applications. Other classical density
studies of spermatozoa prepared in this way to assess their gradient materials, such as sucrose, cesium chloride, Fi-
deoxyribonucleic acid (DNA) stability and chromatin coll, and Metrizamide, failed because dense solutions
damage. were hypertonic and highly viscous.
Adherence Methods Nycodenz, which is based on the iodinated cyclic hy-
drocarbon iohexol, was found to be useful in making den-
These methods should not be directly deleterious to sper-
sity gradients suitable for human sperm preparation (Gel-
matozoa as long as they do not include a prewash step,
lert-Mortimer et al, 1988) and is the basis of OptiPrep
although methods that employ glass fibers (fragments of
(Nycomed Pharma, Oslo, Norway). A dimeric form, io-
which can contaminate the final product) should be care-
dixanol, is also used to make density gradients for euthe-
fully considered before they are accepted for clinical ap-
plications such as preparing spermatozoa for IUI. Only rian, including human, spermatozoa (Accudenz, Nycomed
limited studies of sperm selection and the functional com- Pharma), which appear to perform quite well (Sbracia et
petence of the prepared populations are available that al, 1996; Smith et al, 1997), although its osmotic activity
compare adherence methods with other techniques such requires that the medium in which it is prepared have a
as direct swim-up from semen (DSUS) and density gra- different ionic composition to the usual media used for
dients, although the second generation of SpermPrep col- sperm preparation and is known to support capacitation
umns (SpermPrep, ZBL, Lexington, Ky), which do not and fertilization in vitro.
require that the semen be prewashed, have been reported The plant-derived molecule arabinogalactan has also
to provide poorer yields (Smith et al, 1995). been used to prepare density gradients for human sper-
matozoa under the trade name IsoCare (InVitroCare, San
Density Gradients Diego, Calif), although this author is not aware of any
Early studies (Beatty, 1964) reported that although col- published studies that compare the resulting sperm pop-
loidal silica allowed isopycnic separation of spermatozoa, ulations with other techniques for sperm function or clin-
the fertility of these cells exhibited problems, and it was ical IUI or IVF.
only with the advent of modified colloidal silica in the Although Percoll was based upon colloidal silica coat-
late 1970s that this method became useful. Coating silica ed with PVP, the replacements, PureSperm (Nidacon In-
particles with polyvinylpyrrolidone (PVP) was introduced ternational AB, Göteborg, Sweden) and ISolate (Irvine
by Håkan Pertoft (Pertoft et al, 1978) and commercialized Scientific, Santa Ana, Calif) use colloidal silica with co-
as Percoll (Pertoft’s Colloid), which was used to make valently bound silane molecules (silanized silica). Con-
density gradients for sperm preparations in the early sequently, the characteristics of these colloids should be
1980s. This was a major innovation that to a large extent at least the same as for equivalent Percoll preparations
dominated clinical and experimental human sperm prep-
(Perez et al, 1997), and their clinical utility seems to be
aration until Percoll was withdrawn from clinical use by
at least as good, if not better, than Percoll (Chen and
its manufacturer (Pharmacia Biotech, Uppsala, Sweden)
Bongso, 1999). The major difference is that whereas Per-
in 1996 (additional information on this appears later).
coll was sold as colloidal silica in a weak inorganic buffer
Several intrinsic properties of colloidal silica made Per-
coll (and its recent replacements, PureSperm and ISolate) solution (and was typically used in conjunction with a
ideal for preparing density gradients for selecting human 10⫻ buffer to make ‘‘isotonic’’ 90% vol/vol Percoll),
spermatozoa. First, as a mineral substance, it has no os- PureSperm and ISolate are prepared in what is effectively
motic effect when added to culture medium; second, it an isotonic, ready-to-use culture medium. There are now
allows high-density (high specific gravity) media to be many other products based upon silanized silica on the
prepared, which is important because normal, mature eu- market, and these are likely to be third-party products
therian spermatozoa are dense cells; and third, being a based upon similar raw materials that are used to make
colloid rather than a solution, it has low viscosity and PureSperm and ISolate. An example of these products is
thus does not retard sperm cell sedimentation. Although the Enhance cell isolation product from Conception Tech-
some authors reported the use of continuous Percoll gra- nologies (San Diego, Calif).
 andr 21 316 Mp 360
File # 16em

360 Journal of Andrology · May/June 2000

The Percoll Saga dependent legal opinion on the need to use ART-specific
In October 1996, many laboratories worldwide received a products. In his article, Morroll stated that, ‘‘if a manu-
letter from Pharmacia Biotech that declared that Percoll facturer states that its products may be used only in one
was not to be used for clinical applications and that its use way (e.g., for research purposes only), deviations from
for these purposes was to be prohibited, effective January this are entirely at the risk of the user,’’ and Morroll con-
cluded, ‘‘should one choose to use such a product in an
1, 1997. The letter from the manufacturer stated, ‘‘Phar-
unauthorised manner and a problem arose due to its use,
macia Biotech recognises the ethical and legal obligations
any defence would be non-existent.’’ Morroll also stated,
which compel us to restrict the use of Percoll to RE-
‘‘It may even be argued that one is guilty of a breach of
SEARCH PURPOSES ONLY and to take measures to en-
duty of care. In addition, the manufacturer may opt to
sure it is not used for the isolation of cells which will be
take legal action if its product is used in an unauthorised
subsequently used for clinical purposes in humans.’’ This
way. . . .if a non-ART product is used and is linked with
indicated that the manufacturer was not specifically tar-
a death, this could in theory lead to criminal proceedings,
geting ART laboratories, but the warning was meant for
probably on the grounds of unlawful killing but possibly
anyone who was isolating cells that would be used for
manslaughter if considered reckless.’’ Hence, the legal
clinical purposes for humans, including stem cell therapy.
opinion in the United Kingdom even 3 years ago recog-
This caused quite a furor among Internet special inter-
nized that the onus was on ART clinics and researchers
est groups, such as Androlog ([email protected].
to use ART-specific products exclusively, although an
edu), ARTlog ([email protected]), and EmbryoMail
ACE survey at the time revealed that only 2 clinics in the
([email protected]). Many researchers
United Kingdom had implemented such a policy, even
asked why the step had been taken; in particular, many
though several had indicated that they were considering
believed that because they had used Percoll ‘‘safely’’ for
changing their current policy.
so many years, they could continue to use it regardless
In 1996, Sbracia et al expressed concern that Percoll
of any pronouncement by Pharmacia Biotech, especially
was not approved by FDA and that the manufacturer had
if they bought their Percoll supply from another company,
not intended the product to be marketed for sperm prep-
such as Sigma Chemical Company (St Louis, Mo). How-
aration. Near this time, FDA was preparing to regulate all
ever, because Pharmacia Biotech is the sole manufacturer ART products, and assurances of raw-product purity and
of Percoll, companies such as Sigma were buying it in suitability would have been required for an FDA 510(k)
bulk and repackaging it. There was also concern that submission. As the debate continued, I ventured an opin-
many companies were making ready-to-use gradient kits ion on EmbryoMail 732 (April 2, 1998) that Pharmacia
based on Percoll; yet such kits were all marked ‘‘For In- Biotech’s prohibition could have come about because
Vitro Use Only’’ or ‘‘For Research Use Only,’’ and none third-party companies that assembled Percoll kits were
were sold for the specific application of sperm prepara- seeking assurances that the raw materials in their prepa-
tions for use in human clinical ART. Furthermore, it was rations were pure and suitable. This might have rung
widely recognized that some batches of Percoll contained alarm bells at Pharmacia Biotech. Certainly this was a
high levels of endotoxin contamination, substantially likely scenario, and because Percoll was manufactured
higher than those permissible for in vivo administration, primarily as a research-grade product and the ART market
which could adversely affect sperm survival or develop- represented only a tiny proportion of the total Percoll
ment of fertilized oocytes. The situation was particularly market, there would have been no sound commercial
obscure in the United States because at that time, the U.S. grounds for Pharmacia Biotech to upgrade its manufac-
Government did not recognize IVF. At the same time, turing processes to those that would have been essential
although the U.S. Food and Drug Administration (FDA) to supply a medical product. In addition to these technical
licensed products for use in IUI, it did not license prod- and practical issues, the cost of Percoll would have sky-
ucts that mentioned IVF and, hence, these descriptions rocketed for non-ART users and possibly destroyed that
were generally seen as a labeling tactic. Many workers segment of Pharmica Biotech’s market.
claimed that Pharmacia Biotech did not have the right to I have heard references to ‘‘a report from Europe’’ that
stop them from using Percoll in this way; obviously, this PVP might cause genetic abnormalities in embryos, and
assertion was misguided and incorrect, and anyone who whereas this has been associated with the use of PVP to
continued to use Percoll after the prohibition date ran immobilize spermatozoa during an intracytoplasmic
commercial and medico-legal risks. sperm injection (ICSI) procedure, its use in coating the
The July 1997 issue of The Embryologist, the newslet- silica particles in Percoll (in addition to the free PVP that
ter of the British Association of Clinical Embryologists Percoll was known to contain) could have caused Phar-
(ACE), contained an article by David Morroll, a member macia Biotech, FDA, or both to raise concerns regarding
of the ACE Executive Committee, who reviewed an in- any use of Percoll for ART.
Mortimer · Sperm Preparation Methods 361

The study at the root of this ‘‘report’’ is unknown, but bility issues must be huge, and surely no one wants to be
it may have been the paper by Ray and colleagues (1995) part of a possible criminal litigation.
that included 2 confused references, 1 to an earlier paper
by Fishel and colleagues (1993) on the potential dangers
Medium or Buffer?
of microassisted fertilization and another to an abstract by Because the centrifugation step is performed in air rather
Bras and colleagues (1994), which said that ‘‘some’’ PVP than in a CO2-enriched atmosphere, it is recommended
solutions (from anonymous suppliers) were toxic and that density gradients (and swim-ups) use a HEPES-buff-
caused failed embryonic development after injection into ered medium (often referred to as a sperm wash buffer)
mouse oocytes. However, the paper by Fishel and col- rather than a bicarbonate-based medium (a sperm wash
leagues made no mention of PVP, and besides, any as- medium). However, because sperm capacitation requires
sociation between a particular commercial PVP product the presence of significant concentrations of bicarbonate
and abnormal embryo development is an entirely separate ions, the washing step and final resuspension must be
issue! Moreover, the study by Ray and colleagues found made into a medium; otherwise, in vitro capacitation and
no evidence that PVP or methyl cellulose caused DNA hence fertilization can be compromised. For IUI prepa-
lesions and concluded that their data provided reassuring rations, a buffer can be employed because it will be great-
evidence for the use of those products in sperm injection ly diluted after insemination, and because if the sperma-
procedures. tozoa undergo capacitation in vitro during prolonged in-
A third consideration is Percoll’s variable and some- cubation before insemination, their hyperactivated motil-
times high endotoxin levels. This can be tested for by end ity could compromise their ability to traverse the
users, and only ‘‘safe’’ batches can be used in clinical uterotubal junction (Shalgi et al, 1992; Mortimer, 1997),
procedures. However, this would continue to make Percoll but if insemination is to be performed soon after sperm
an unreliable raw material according to good manufac- preparation, then a medium can be safely used.
turing practices, standards by which all medical device Should We Include Antioxidants in Sperm Preparation
manufacturers must operate, and would elevate a clinic’s Media?
costs by purchasing unusable batches of the material. Because spermatozoa may be exposed to potentially haz-
Percoll has been used, apparently safely, for many ardous effects of ROS during preparation, several workers
years in many ART laboratories (although it will remain have suggested including antioxidant protection (eg, glu-
a mystery as to whether any poor or failed fertilization tathione) in sperm preparation media formulations. Al-
cases or poor or failed embryonic development might though a slightly improved yield supports this in prelim-
have been caused by undetected, elevated endotoxin lev- inary evidence (Parinaud et al, 1997), more extensive
els in Percoll). But the next-generation products that are studies will be needed before the value of this concept
based on silanized colloidal silica particles must be man- can be established.
ufactured according to good manufacturing practices for
clinical use, for which Percoll was never intended. In ad-
dition, new products should receive regulatory approval Assessing the Yield of a Sperm
from the FDA for at least IUI use, and preferably unre- Preparation Method
stricted ART, as well as clearance for use as a medical
device by other regulatory authorities, including the Eu- The success or applicability of a sperm-washing method
ropean Medical Device Directive. The attendant increased can be considered in terms of either the absolute or rel-
cost of such products and the task of obtaining interna- ative yield of motile spermatozoa that one obtains at the
tional regulatory approvals is attributable to the need to end of the technique. Usually, progressive motility is used
follow these higher manufacturing standards, but the cost for this purpose because nonprogressive spermatozoa are
should be only about $5 per procedure (presuming a fair unlikely to be potentially functional (except perhaps for
price is set by distributors), which is an insignificant cost ICSI).
within the IUI cycle and absolutely trivial in view of the Relative yield is the proportion of progressively motile
cost of IVF. Regardless of budget squeezes on ostensibly spermatozoa submitted to a preparative procedure that are
commercial grounds, everyone involved in medical care present in the final preparation. It is calculated the fol-
must strive for best practice, and the replacement of Per- lowing way:
coll by products such as PureSperm or ISolate has to be
yield (%) ⫽ (v ⫻ c ⫻ pm%)/(V ⫻ C ⫻ PM%) ⫻ 100,
viewed in this light. And those laboratories that still use
Percoll must question their motives and balance the small where v is the final preparation volume, V is the volume
savings they are achieving against the risks of using an of semen used, c is the sperm concentration in the final
unregistered product whose manufacturer has declared it preparation, pm% is the prepared sperm population pro-
to be unsuitable for ART applications. The potential lia- gressive motility, C is the semen sperm concentration, and
 andr 21 316 Mp 362
File # 16em

362 Journal of Andrology · May/June 2000

PM% is the progressive motility of spermatozoa in the Several laboratories reported the use of hypertonic Per-
semen. coll gradients that were designed to minimize such os-
Absolute yield is the total number of progressively mo- motic shocks (Velez de la Calle, 1991; Mortimer, 1994b),
tile spermatozoa that can be obtained if the whole ejac- and whereas postrecovery sperm behavior was more nor-
ulate is used, although an allowance is usually made for mal, the yield was more variable. This was because of
the aliquots needed to perform a standard semen analysis the basis upon which density gradients operate and the
(eg, 0.3 mL). Yield quality is, however, of vital impor- highly variable morphology of spermatozoa in the ejac-
tance, especially if the product is to be used to create ulates of different men. Because density gradients operate
embryos for clinical purposes. on the basis of cells’ specific gravity, with centrifugation
causing them to move down the gradient to their isopyc-
Modifying the Yield by Altering the Colloid nic point, the presence of variable amounts of retained
Concentration cytoplasm in the spermatozoa will cause them to have
This section considers how the colloid concentration and different densities. Normal spermatozoa with no retained
osmolarity of the gradient influence yield. Because much cytoplasm are very dense because of their condensed
of this research was performed several years ago, it em- chromatin, and they reach the bottom of the centrifuge
ployed Percoll and has been presented on this basis. This tube because their density is slightly greater than that of
must not be taken as any suggestion that Percoll should the lower or lowest layer (about 80% Percoll in this dis-
be used for any clinical application nowadays. In simple cussion). Therefore, because research donors are selected
terms, the purchased Percoll product can be taken as on the basis of their excellent semen characteristics, they
100% colloid and equivalent to, for example, 100% should have the highest proportions of very dense sper-
PureSperm for any application of these principles. matozoa and will provide good yields, even with a 95%
The earliest Percoll method we used had a true 95% lower layer—and semen specimens from patients who are
vol/vol concentration of Percoll in its bottom layer (Drav- infertile are likely to perform substantially less well. Re-
land and Mortimer, 1985), and although it provided ex- ducing the lower layer to 72% Percoll allows less dense
cellent yields when used with research donors, in clinical spermatozoa to reach the bottom of the tube, which re-
specimens from an infertility clinic, the yield was much sults in an increase in total yield, but only because sper-
lower, sometimes approaching zero. Empirical studies matozoa with some retained cytoplasm are recovered.
demonstrated that optimum clinical yields could be ob- For the aforementioned reasons and because the mod-
tained with about 80% vol/vol in the lower Percoll layer, ern products such as PureSperm and ISolate are sold as
and whereas 72% would often result in at least slight isotonic colloidal preparations that eliminate the need for
increases in the absolute yield, it seemed to be at the the 10⫻ mixing step, all workers are urged to consider
expense of recovering more abnormal spermatozoa (Mor- how they report the composition of their gradients, to use
timer, 1994b). Consequently, we always recommend the scientifically correct descriptions, and thus to avoid the
use of a lower layer of 81% Percoll to obtain optimum confusion that can result in the use of inappropriate gra-
yields (obtained as 90% of the 90% isotonic Percoll). This dient formulations (eg, the protocol for a Percoll-based
procedure was essential when working with Percoll be- gradient published in the third edition of the WHO lab-
cause the osmolarity of true 100% (ie, the stock product oratory manual [World Health Organization, 1992] that
as supplied in the bottle from Pharmacia or Sigma is had a 72% lower layer; that is, 80% of 90%).
about 17 mOsm, so it was mixed as 9 ⫹ 1 with a 10⫻ Another important point is that cryopreserved sper-
medium so that if the medium was, say, based upon matozoa are in a highly hypertonic medium and hence
Quinn’s HTF at 285 mOsm, the isotonic Percoll would will suffer extreme osmotic shock upon entering the up-
have an osmolarity of 258 mOsm; that is, ([9 ⫻ 285] ⫹ per layer of a density gradient. This is the explanation for
17)/10. However, the incorrect practice by some authors the obligate requirement of a slow dilution of cryopre-
of renaming the 90% isotonic Percoll preparation as a served semen after thawing with a large volume of culture
100% preparation has created enormous confusion. These medium so as to bring the osmolarity of the sperm sus-
osmotic shifts were the generally accepted explanation of pension closer to that of the gradient before beginning the
why spermatozoa recovered from a Percoll gradient and first centrifugation step (Ford et al, 1992).
then washed in fresh medium (at 285 mOsm) seemed to
swim with jerky movements for the first few minutes after Modifying the Yield by Altering the Centrifugation Speed
final resuspension while they were recovering from the All early reports of Percoll gradients employed an initial
combined osmotic shock of going from about 340 mOsm centrifugation step through Percoll at 300 ⫻ g, which em-
in the seminal plasma to 258 mOsm in the lower gradient pirically had been found to provide optimum sperm re-
layer and back to 285 mOsm in the culture medium, in covery (Dravland and Mortimer, 1985). Increasing the
addition to the general effect of centrifugation. centrifugation speed, time, or both could sometimes in-
Mortimer · Sperm Preparation Methods 363

crease the number of spermatozoa recovered, but not nec- separation of spermatozoa on density gradients based on
essarily substantially—or usefully, in terms of the quality colloidal silica are true isopycnic separation methods but
of the extra spermatozoa recovered. Over many years of are perhaps unusual in that they cannot achieve sufficiently
experience we have found there to be no real benefit in high densities (at least using currently available colloidal
altering this initial centrifugation step from 20 minutes at silica materials) to separate the spermatozoa as a discrete
300 ⫻ g, and we continue to make this recommendation. band above the bottom of the tube.
Because the purchased Percoll product (100% colloid) is
equivalent to 100% PureSperm, the same optimized per-
formance can be expected using the same centrifugation Sperm Selection
conditions with this product, and because ISolate is de-
scribed as being the same as 100% Percoll, one would Sperm Motility and Morphology
expect the same for that product as well (although I have Any method based upon sperm migration will certainly
never used ISolate). produce a sperm population that is selected for improved
The subsequent wash centrifugation step should be suf- sperm morphology. However, this has long been known
ficient to recover the great majority of the previously pel- to be based upon the differential distribution of midpiece
leted spermatozoa without exposing them to excessive cen- and tail defects among spermatozoa with normal and ab-
trifugal force. In practice, this has been achieved at 500 ⫻ normal head morphology (Mortimer et al, 1982), so that
g for 10 minutes; again, longer centrifugation does not pro- applying a selection pressure based upon motility will
vide any significant benefit, although a slightly shorter time create a concomitant improvement in overall sperm mor-
of, say, 6 minutes can be used if the speed is increased to phology. This process is equivalent to that occurring at
600 ⫻ g. Lower speeds can be used, but there is an in- the level of penetration of the cervical mucus in vivo and
creased risk of losing some of the spermatozoa in the dis- reflects a natural process of sperm selection. However, it
carded supernatant, and whether there are fewer good sper- has also long been established that this process of sperm
matozoa with lower densities is unknown. Under no cir- phenotypic selection does not provide a general selection
cumstances should the speed exceed 800 ⫻ g. for spermatozoa with normal genotype, except perhaps
Reports must not describe their methods using revo- for reductions in diploid spermatozoa that have larger
lutions per minute because they are highly influenced by heads (Carothers and Beatty, 1975).
rotational radius and the centrifugal forces discussed in
the previous paragraph are actually gmax values that were Sperm Morphology (Phenotype) and Genotype
calculated for what the spermatozoa would experience at There are few instances in which sperm morphology as-
the bottom of the centrifuge tube. The formula describing sessed by light microscopy is related to the genetic con-
the relationship between rotation speed, rotational radius, tent of the spermatozoa. Abnormally small and large
and centrifugal force is as follows: sperm heads are often associated with aneuploidy and
perhaps diploidy (Carothers and Beatty, 1975) and the t
g ⫽ 0.0000112 ⫻ r ⫻ N2, or N ⫽ 兹(g/[0.0000112 ⫻ r]),
loci in mice cause the production of abnormal sperma-
where g ⫽ the maximum centrifugal force achieved at the tozoa (Olds-Clarke, 1990). More specific associations also
bottom of the tube, r ⫽ the rotational radius (in centi- exist; for example, the total lack of sperm motility in men
meters), and N ⫽ revolutions per minute. with Kartagener (or immotile cilia) syndrome and the de-
fect known as globozoospermia, although their patterns
Do Spermatozoa Swim Through the Gradients? of inheritance remain unknown, notably because hitherto,
Several workers have commented that spermatozoa swim men with these defects were sterile. Reports of morpho-
through the density gradients and that the centrifugal force metric differences between X- and Y-bearing spermato-
helps align them in a downward direction, a concept that zoa—and futile attempts to separate the 2 populations of
may be based on some reports of using simple migration spermatozoa based upon them (Gledhill, 1988; Martin,
through density gradients under unit gravity. Additional 1994)—are numerous, and whereas specific size differ-
centrifugal force will certainly cause human spermatozoa ences seem to exist (Cui, 1997), they are too small to
to align in a head-down orientation (rabbit spermatozoa do have practical consequence (although they do confirm the
this under unit gravity because of the larger size of their association between sperm head size and total chromo-
heads; Branham, 1969; Mortimer, 1979), but the relative some content).
contributions of sperm motility and settling under a greatly Several authors have expressed concern over the use of
increased gravitational force are clearly unequal—and fur- ICSI, in which spermatozoa do not go through the same
thermore, the less dense spermatozoa never reach the bot- selection process as they would in vivo and hence may
tom of the gradient but remain at their isopycnic level. So contribute to an increased prevalence of genetic anoma-
the conclusion, on the basis of simple physics, is that the lies in the resulting offspring. The grounds for this sup-
 andr 21 316 Mp 364
File # 16em

364 Journal of Andrology · May/June 2000

posed relationship are, however, generally baseless at the matozoa with a lower incidence of DNA nicks to be re-
light microscopy level. It has been shown that morpho- covered (Sakkas, in press). Other studies using SCSA
logically abnormal spermatozoa carry normal karyotypes have shown that swim-up (Spanò et al, 1999) and glass
(Martin and Rademaker, 1988) and can produce normal wool filtration (Larson et al, 1999) can select spermatozoa
offspring (Burruel et al, 1996), although it has also re- with better chromatin stability. However, because SCSA
cently been reported that human and mouse spermatozoa assesses the susceptibility of sperm DNA to in vitro acid-
with abnormal head shape can have an increased inci- or temperature-induced damage rather than the actual ex-
dence of structural chromosomal aberrations (Lee et al, istence of nicks in the sperm DNA, the relative value of
1996; Kishikawa et al, 1999). Nevertheless, we know that the 2 assays, and hence their clinical significance, remain
under normal in vivo conception conditions in fertile cou- to be determined.
ples, many genetically abnormal embryos are produced The vital importance of these observations is that,
and transmitted via the spermatozoa, many of which are whereas many ART laboratories employ density gradient
caused by aneuploidy and even triploidy. Consequently, preparation methods for their IVF spermatozoa, many use
such calls for abandoning ICSI because it does not prop- simple washing for ICSI because the risk of ROS-induced
erly select spermatozoa are alarmist and, on the basis of sperm dysfunction arising from simple centrifugal wash-
our current knowledge, largely unfounded. ing is considered unimportant because sperm function is
This subject is a major area of current research and generally accepted as irrelevant to fertilization by ICSI.
space precludes a more detailed review in this article. However, the Brussels group that originally developed the
Interested readers will find additional discussions in the ICSI technique employ density gradients (originally Per-
proceedings of the second Collioure conference Genetics coll, now PureSperm; see Verheyen et al, 1999). But the
of Human Male Fertility (Barratt et al, 1997) and else- sequel to this is that the men who are considered to need
where (eg, Mortimer, 2000). ICSI will be more likely to have increased proportions of
abnormal spermatozoa, and hence they will also be the
Free Radicals and Sperm Chromatin Damage most susceptible to ROS-induced sperm DNA damage.
The deleterious effect of free radicals or ROS upon sperm Because it has been well established that ROS can cause
function and their role in the etiology of male infertility substantial degradation of the sperm DNA that does not
was originally described in John Aitken’s laboratory in Ed- necessarily affect their fertilizing ability (Aitken et al,
inburgh (Aitken and Clarkson, 1987). Soon afterward, the 1998), spermatozoa prepared by simple washing will def-
significance of ROS generation during sperm preparation initely be at a much greater risk of contributing a defec-
in vitro was described (Aitken and Clarkson, 1988) and the tive genome to the embryo and could underlie the in-
existence of widespread evidence for the detrimental influ- creased developmental failure of ICSI-derived embryos
ence that this could have upon sperm function tests and after the 8-cell stage when the embryonic genome is ac-
IVF was collated (Mortimer, 1991). ROS are generated tivated (Shoukir et al, 1998).
both by leukocytes present in semen and spermatozoa
(Krausz et al, 1992; Aitken, 1995; Whittington and Ford,
1999). However, only those spermatozoa with excess re- Conclusion
tained spermatid cytoplasm generate ROS (Aitken et al,
1994; Aitken, 1995; Huszar et al, 1998, 1999), which il- According to our current knowledge, colloidal silica den-
lustrates the highly beneficial selection of spermatozoa that sity gradients must be considered the most appropriate
density gradients confer by eliminating these less dense and generally applicable clinical sperm preparation tech-
spermatozoa from the final preparation. ROS affect not nique. If other methods, such as some of the adherence-
only the sperm plasma membrane by causing phospholipid based products, are able to separate spermatozoa with
peroxidation, and hence decreased membrane fluidity and comparably reduced levels of DNA damage, then this
impaired sperm function, but also the sperm DNA by caus- must be demonstrated through independent scientific
ing strand breaks that can be revealed by various tests of study. Indeed, it would seem essential that any future
sperm DNA integrity such as nick translation (Sakkas et studies on the development of new sperm preparation
al, 1997) and the sperm chromatin structure assay (SCSA; methods or that report comparative analyses of various
Evenson, 1999; Evenson et al, 1999). methods must include sperm DNA assessments in order
Of particular importance in the practice of ART is a to have real clinical value. Moreover, because the most
study comparing the incidence of DNA nicks in sper- fundamental guiding principle of medical care is primum
matozoa recovered by simple washing, swim-up from se- non nocere, or first, do no harm, physicians and clinical
men, and colloidal silica–based density gradient separa- scientists who participate in the management of ART pro-
tion (Percoll and PureSperm) in which only the density grams are obligated to avoid techniques that have known
gradient methods permitted selected populations of sper- hazards if other, safer techniques are available. It seems
Mortimer · Sperm Preparation Methods 365

highly unlikely that any couple undergoing ART would Englert Y, Van den Bergh M, Rodesch C, Bertrand E, Biramane J, Le-
choose to save $5 by using a simple sperm-washing tech- greve A. Comparative auto-controlled study between swim-up and
Percoll preparation of fresh semen samples for in-vitro fertilization.
nique instead of density gradients if the true risks, in Hum Reprod. 1992;7:399–402.
terms of an increased chance of embryonic wastage re- Evenson DP. Alterations and damage of sperm chromatin structure and
sulting from an elevated incidence of genetic anomalies early embryonic failure. In: Jansen R, Mortimer D, eds. Towards Re-
in the embryos, were fully explained during the consent productive Certainty: Fertility and Genetics Beyond 1999. Carnforth,
process. It is clearly important that clinical units use only United Kingdom: Parthenon Publishing; 1999.
Evenson DP, Jost LK, Marshall D, Zinaman MJ, Clegg E, Purvis K, de
sperm preparation products that are designed and ap- Angelis P, Claussen OP. Utility of the sperm chromatin structure assay
proved by relevant (ie, local) regulatory authorities for as a diagnostic and prognostic tool in the human fertility clinic. Hum
ART use. Reprod. 1999;14:1039–1049.
Fishel S, Dowell K, Timson J, Green S, Hall J, Klentzeris L. Micro-
assisted fertilization with human gametes. Hum Reprod. 1993;8:
References 1780–1784.
Ford WCL, McLaughlin EA, Prior SM, Rees JM, Wardle PG, Hull MGR.
Aitken RJ. Free radicals, lipid peroxidation and sperm function. Reprod The yield, motility and performance in the hamster egg test of human
Fertil Dev. 1995;7:659–668. spermatozoa prepared from cryopreserved semen by four different
Aitken RJ, Clarkson JS. Cellular basis of defective sperm function and methods. Hum Reprod. 1992;7:654–659.
its association with the genesis of reactive oxygen species by human Gellert-Mortimer ST, Clarke GN, Baker HWG, Hyne RV, Johnston WIH.
spermatozoa. J Reprod Fertil. 1987;81:459–469. Evaluation of Nycodenz and Percoll density gradients for the selection
Aitken RJ, Clarkson JS. Significance of reactive oxygen species and an-
of motile human spermatozoa Fertil Steril. 1988;49:335–441.
tioxidants in defining the efficacy of sperm preparation techniques. J
Gledhill BL. Selection and separation of X- and Y- chromosome-bearing
Androl. 1988;9:367–376.
mammalian spermatozoa. Gamete Res. 1988;20:377–395.
Aitken RJ, Gordon E, Harkiss D, Twigg JP, Milne P, Jennings Z, Irvine
Gorus FK, Pipeleers DG. A rapid method for the fractionation of human
DS. Relative impact of oxidative stress on the functional competence
spermatozoa according to their progressive motility. Fertil Steril.
and genomic integrity of human spermatozoa. Biol Reprod. 1998;59:
1981;35:662–665.
1037–1046.
Guérin JF, Mathieu C, Lornage J, Pinatel MC, Boulieu D. Improvement
Aitken RJ, Krausz C, Buckingham DW. Relationship between the bio-
of survival and fertilizing capacity of human spermatozoa in an IVF
chemical marker for human sperm quality, creatine kinase, and the
programme by selection on discontinuous Percoll gradients. Hum Re-
mechanisms responsible for the generation of reactive oxygen species.
prod. 1989;4:798–804.
Mol Reprod Dev. 1994;39:268–279.
Huszar G, Patrizio P, Vigue L, Willets M, Wilker C, Adhoot D, Johnson
Arcidiacono A, Walt H, Campana A, Balerna M. The use of Percoll
L. Cytoplasmic extrusion and the switch from creatine kinase B to M
gradients for the preparation of subpopulations of human spermato-
isoform are completed by the commencement of epididymal transport
zoa. Int J Androl. 1983;6:433–445.
in human and stallion spermatozoa. J Androl. 1998;19:11–20.
Barratt C, De Jonge C, Mortimer D, Parinaud J, eds. Genetics of Human
Huszar G, Willets M, Corrales M. Hyaluronic acid (SpermSelect) im-
Male Fertility. Paris: EDK; 1997.
proves retention of sperm motility and velocity in normospermic and
Beatty RA. Density gradient media for mammalian spermatozoa. In: Pro-
oligospermic specimens. Fertil Steril. 1990;54:1127–1134.
ceedings of the Fifth International Congress on Animal Reprodtive
Huszar G, Zeyneloglu HB, Vigue L. Cellular maturity and fertilizing
Artificial Insemination. III Part 2. Trento, Italy: 1964;276–281.
potential of sperm populations in natural and assisted reproduction.
Bolton VN, Braude PR. Preparation of human spermatozoa for in vitro
In: Gagnon C, ed. The Male Gamete: From Basic Science to Clinical
fertilization by isopycnic centrifugation on self-generating density
gradients. Arch Androl. 1984;13:167–176. Applications. Vienna, Ill: Cache River Press; 1999.
Branham JM. Movements of free-swimming rabbit spermatozoa. J Re- Jeulin C, Serres C, Jouannet P. The effects of centrifugation, various
prod Fertil. 1969;18:97–105. synthetic media and temperature on the motility and vitality of human
Bras M, Dumoulin JCM, Pieters MHEC, Michaels AHJC, Geraedts JPM, spermatozoa. Reprod Nutr Dev. 1982;22:81–91.
Evers JLH. The use of a mouse zygote quality control system for Kanwar KC, Yanagimachi R, Lopata A. Effects of human seminal plasma
training purposes and toxicity determination in an ICSI programme. on fertilizing capacity of human spermatozoa. Fertil Steril. 1979;31:
Hum Reprod. 1994;9(Suppl 4):23–24. 321–327.
Burruel VR, Yanagimachi R, Whitten WK. Normal mice develop from Katz DF, Morales P, Samuels SJ, Overstreet JW. Mechanisms of filtration
oocytes injected with spermatozoa with grossly misshapen heads. Biol of morphologically abnormal human sperm by cervical mucus. Fertil
Reprod. 1996;55:709–714. Steril. 1990;54:513–516.
Carothers AD, Beatty RA. The recognition and incidence of haploid and Kishikawa H, Tateno H, Yanagimachi R. Chromosome analysis of BALB/
polyploid spermatozoa in man, rabbit and mouse. J Reprod Fertil. c mouse spermatozoa with normal and abnormal head morphology.
1975;44:487–500. Biol Reprod. 1999;61:809–812.
Chan SYW, Tucker MJ. Differential sperm performance as judged by the Krausz C, West K, Buckingham D, Aitken RJ. Development of a tech-
zona-free hamster egg penetration test relative to differing sperm pen- nique for monitoring the contamination of human semen samples with
etration techniques. Hum Reprod. 1992;7:255–260. leukocytes. Fertil Steril. 1992;57:1317–1325.
Chen M-J, Bongso A. Comparative evaluation of two density gradient Kremer J. The in vitro spermatozoal penetration test in fertility investi-
preparations for sperm separation for medically assisted conception. gations. PhD dissertation. University of Groningen, The Netherlands;
Hum Reprod. 1999;14:759–764. 1968.
Cui K-H. Size differences between human X and Y spermatozoa and Larson KL, Brannian JD, Timm BK, Jost L, Evenson DP. Density gra-
prefertilization diagnosis. Mol Hum Reprod. 1997;3:61–67. dient centrifugation and glass wool filtration of semen remove sper-
Dravland JE, Mortimer D. A simple discontinuous Percoll gradient proce- matozoa with damaged chromatin structure. Hum Reprod. 1999;14:
dure for washing human spermatozoa. IRCS Med Sci. 1985;13:16–17. 2015–2019.
 andr 21 316 Mp 366
File # 16em

366 Journal of Andrology · May/June 2000

Lee JD, Kamiguchi Y, Yanagimachi R. Analysis of chromosome consti- Sakkas D, Manicardi GC, Tomlinson M, Mandrioli M, Bizzaro D, Bianchi
tution of human spermatozoa with normal and aberrant head mor- PG, Bianchi U. The use of two dense degraded centrifugation tech-
phologies after injection into mouse oocytes. Hum Reprod. 1996;1: niques and swim-up sperm separation to separate spermatozoa with
1942–1946. chromatin and nuclear DNA anomalies. Human Reprod. In press.
Lessley BA, Garner DL. Isolation of motile spermatozoa by density gra- Sakkas D, Bianchi PG, Manicardi G, Bizzaro D, Bianchi U. Chromatin
dient centrifugation in Percoll. Gamete Res. 1983;49–61. packaging anomalies and DNA damage in human sperm: their pos-
Martin RH. Human sex pre-selection by sperm manipulation. Hum Re- sible implications in the treatment of male factor infertility. In: Barratt
prod. 1994;9:1790–1791. C, De Jonge C, Mortimer D, Parinaud J, eds. Genetics of Human
Martin RH, Rademaker A. The relationship between sperm chromosomal Male Fertility. Paris: EDK; 1997.
abnormalities and sperm morphology in humans. Mutat Res. 1988; Sbracia M, Grasso J, Sayme N, Stronk J, Huszar G. Hyaluronic acid
207:159–164. substantially increases the retention of motility in cryopreserved/
Morales P, Vantman D, Barros C, Vigil P. Human spermatozoa selected thawed human spermatozoa. Hum Reprod. 1997;12:1949–1954.
by Percoll gradient or swim-up are equally capable of binding to the Sbracia M, Sayme N, Grasso J, Vigue L, Huszar G. Sperm function and
human zona pellucida and undergoing the acrosome reaction. Hum choice of preparation media: comparison of Percoll and Accudenz
Reprod. 1991;6:401–404.
discontinuous density gradients. J Androl. 1996;17:61–67.
Mortimer D. Differential motility of diploid rabbit spermatozoa. Arch
Shalgi R, Smith TT, Yanagimachi R. A quantitative comparison of the
Androl. 1979;2:41–47.
passage of capacitated and uncapacitated hamster spermatozoa
Mortimer D. Sperm preparation techniques and iatrogenic failures of in-
through the uterotubal junction. Biol Reprod. 1992;46:419–424.
vitro fertilization. Hum Reprod. 1991;6:173–176.
Shoukir Y, Chardonnens D, Campana A, Sakkas D. Blastocyst develop-
Mortimer D. Practical Laboratory Andrology. New York: Oxford Uni-
ment from supernumerary embryos after intracytoplasmic sperm in-
versity Press; 1994a.
jection: a paternal influence? Hum Reprod. 1998;13:1632–1637.
Mortimer D. Sperm recovery techniques to maximize fertilizing capacity.
Reprod Fertil Dev. 1994b;6:25–31. Smith S, Hosid S, Scott L. Use of postseparation sperm parameters to
Mortimer D. Sperm transport in the female genital tract. In: Grudzinskas determine the method of choice for sperm preparation for assisted
JG, Yovich, JL, eds. Cambridge Reviews in Human Reproduction, Vol reproductive technology. Fert Steril. 1995;63:591–597.
2: Gametes—The Spermatozoon. Cambridge, United Kingdom: Cam- Smith TT, Byers M, Kaftani D, Whitford W. The use of iodixanol as a
bridge University Press; 1995. density gradient material for separating human sperm from semen.
Mortimer D. The future of male infertility and ART. Hum Reprod. In Arch Androl. 1997;38:223–230.
press. Spanò M, Cordelli E, Leter G, Lombardo F, Lenzi A, Gandini L. Nuclear
Mortimer D, Leslie EE, Kelly RW, Templeton AA. Morphological selec- chromatin variations in human spermatozoa undergoing swim-up and
tion of human spermatozoa in vivo and in vitro. J Reprod Fertil. cryopreservation evaluated by flow cytometric sperm chromatin struc-
1982;64:391–399. ture assay. Mol Hum Reprod. 1999;5:29–37.
Mortimer D, Mortimer ST. Methods of sperm preparation for assisted Tomlinson M, Manicardi GC, Mandrioli M, Bizzaro D, Bianchi PG, Bian-
reproduction. Ann Acad Med Singapore. 1992;21:517–524. chi U, Turner J, Sakkas D. The use of PureSperm, mini-Percoll and
Mortimer ST. A critical review of the physiological importance and anal- swim-up sperm preparation techniques to separate spermatozoa with
ysis of sperm movement in mammals. Hum Reprod Update. 1997;3: chromatin and nuclear DNA anomalies. Andrology in the Nineties,
403–439. Cape Town, South Africa, March 1999. Abstract P-1.
Mortimer ST, Swan MA, Mortimer D. Effect of seminal plasma on ca- Velez de la Calle JF. Human spermatozoa selection in improved discon-
pacitation and hyperactivation in human spermatozoa. Hum Reprod. tinuous Percoll gradients. Fertil Steril. 1991;56:737–742.
1998;13:2139–2146. Verheyen G, Tournaye H, Staessen C, De Vos A, Vandervorst M, Van
Olds-Clarke P. Genetic analysis of sperm motility. In: Gagnon C, ed. Steirteghem A. Controlled comparison of conventional in-vitro fertil-
Controls of Sperm Motility. Boca Raton, Fla: CRC Press; 1990. ization and intracytoplasmic sperm injection in patients with asthen-
Parinaud J, Le Lannou D, Vieitez G, Griveau J-F, Milhet P, Richoilley G. ozoospermia. Hum Reprod. 1999;14:2313–2319.
Enhancement of motility by treating spermatozoa with an antioxidant
Whittington K, Ford WCL. Relative contribution of leukocytes and of
solution (Sperm-Fit) following ejaculation. Hum Reprod. 1997;12:
spermatozoa to reactive oxygen species production in human sperm
2434–2436.
suspensions. Int J Androl. 1999;22:229–235.
Perez SM, Chan PJ, Patton WC, King A. Silane-coated silica particle
Wikland M, Wik O, Steen Y, Qvist K, Söderlund B, Janson PO. A self-
colloid processing of human sperm. J Assist Reprod Genet. 1997;14:
migration method for preparation of sperm for in-vitro fertilization.
388–393.
Pertoft H, Laurent TC, Laas T. Density gradients prepared from colloidal Hum Reprod. 1987;2:191–195.
silica particles coated by polyvinyl-pyrrolidone (Percoll). Anal Bioch- World Health Organization. WHO Laboratory Manual for the Analysis of
em. 1978;88:271–277. Human Semen and Sperm-Cervical Mucus Interaction. Third ed.
Ray BD, Howell RT, McDermott A, Hull MGR. Testing the mutagenic Cambridge, United Kingdom: Cambridge University Press.
potential of polyvinylpyrrolidone and methyl cellulose by sister chro- Yanagimachi R. Mammalian fertilization. In: Knobil E, Neill JD, eds.
matid exchange analysis prior to use in intracytoplasmic sperm injec- The Physiology of Reproduction. 2nd ed. New York: Raven Press;
tion procedures. Hum Reprod. 1995;10:436–438. 1994.
Rogers BJ, Perreault SD, Bentwood BJ, McCarville C, Hale RW, Sod- Zimmerman ER, Robertson KR, Kim BS, Drobnis EZ, Nakajima ST.
erdahl DW. Variability in the human-hamster in vitro assay for fertility Semen preparation with the Sperm Select system versus a washing
evaluation. Fertil Steril. 1983;39:204–211. technique. Fertil Steril. 1994;61:269–275.

Andrology Lab Corner welcomes the submission of unsolicited manuscripts, requested reviews, and articles in a debate format.
Manuscripts will be reviewed and edited by the Section Editor. All submissions should be sent to the Journal of Andrology Editorial
Office. Letters to the editor in response to articles as well as suggested topics for future issues are encouraged.