Theriogenology 59 (2003) 1241±1255

In¯uence of extender, temperature, and addition

of glycerol on post-thaw sperm parameters
in ram semen
Jorge Gila,b, Nils Lundeheimb,c, Lennart SoÈderquista,b,
Heriberto RodrõÂguez-MartõÂneza,b,*
a
Department of Obstetrics and Gynecology, Faculty of Veterinary Medicine,
Swedish University of Agricultural Sciences (SLU), UllsvaÈgen 14C,
Box 7039, Uppsala SE-750 07, Sweden
b
Center for Reproductive Biology in Uppsala (CRU),
Swedish University of Agricultural Sciences (SLU), Uppsala, Sweden
c
Department of Animal Breeding and Genetics, Faculty of Agriculture,
Swedish University of Agricultural Sciences (SLU), Uppsala, Sweden
Received 13 September 2001; accepted 10 June 2002

Abstract

Using a two-step extension methodology, two experiments were conducted using a split-sample
design to compare the effect on post-thaw ram sperm parameters of a milk-based extender
(Experiment 1) containing four different egg yolk concentrations (5% [M5], 10% [M10], 15%
[M15], and 20% [M20]), and a commercially available extender (Bioexcell1; IMV, L'Aigle,
France) free from additives of animal origin, containing two different ®nal glycerol concentrations
(3.2% [B] and 6.4% [BB]) (Experiment 2). In both experiments, glycerol was added either at 5 8C or
at 15 8C together with the second fraction of each extender. The sperm characteristics assessed were
motility (measured subjectively [SM] and by means of cell motion analysis (CASA)), membrane
integrity (SYBR-14/PI), and capacitation status (chlortetracycline (CTC)/EthD-1). Results of
Experiment 1 showed no signi®cant positive effect of increasing the concentration of egg yolk
above 10% on post-thaw motility, membrane integrity, or induction of sperm capacitation-like
changes. In Experiment 2, Bioexcell1 (BB) yielded similar post-thaw results as did the milk
extender (control). In both experiments, post-thaw sperm parameters were better preserved when
glycerol was added at 5 8C, although the results were not always statistically signi®cant for
all variables studied. In conclusion, when using milk-based extenders for freezing ram semen, low
(5±10%) concentrations of egg yolk and the addition of glycerol at 5 8C are recommended.
Furthermore, the results indicate that when freezing ram semen, Bioexcell1 containing 6.4%

*
Corresponding author. Tel.: ‡46-186-72172; fax: ‡46-186-73545.
E-mail address: heriberto.rodriguez@og.slu.se (H. RodrõÂguez-MartõÂnez).

0093-691X/02/$ ± see front matter # 2002 Elsevier Science Inc. All rights reserved.
PII: S 0 0 9 3 - 6 9 1 X ( 0 2 ) 0 1 1 7 7 - 9
1242 J. Gil et al. / Theriogenology 59 (2003) 1241±1255

glycerol may be used as an alternative extender to the conventional milk extender containing 5%
egg yolk.
# 2002 Elsevier Science Inc. All rights reserved.

Keywords: Capacitation; Extender; Post-thaw; Ram; Semen; Viability

1. Introduction

The sheep industry is interested in improving long-term semen preservation, particularly

for use in cervical arti®cial insemination (AI), an approach that is more practical than
intrauterine deposition of semen [1±3]. However, the relatively low success rate of cervical
AI with frozen semen in sheep has limited a wider application of this technique, calling for
an improvement of the insemination technique itself, and of the survival rate of the frozen±
thawed semen. Suitable extenders to preserve an adequate number of spermatozoa with all
the attributes needed to overcome the cervical barrier and to achieve fertilization post-
thawing, are needed to achieve these goals [4±7].
In Norway and Sweden, the routine protocol for freezing ram semen consists of a two-
step extension method using a milk-based extender. First, the semen is diluted with an
extender without glycerol (fraction 1 (f/1)), and cooled to 5 8C, then further diluted 1:1
with an extender (fraction 2 (f/2)) containing a glycerol concentration of 14%, resulting in a
®nal concentration of 7%. It is equilibrated for 2 h, centrifuged to re-concentrate the sperm
concentration to 200  106 spermatozoa/AI dose, packed in 0.25 ml mini-straws, and
®nally frozen in LN2 [8].
We have previously studied in vitro as well as in vivo the effect of an adjustment of the
extension rate of the semen as an alternative technique that would exclude the centrifuga-
tion step, thus simplifying the processing routines and avoiding the stress of centrifugation
on spermatozoa [9,10]. Our results indicated that the quality of the AI dose was maintained,
and that despite the prefreezing semen/extender ratio being below that known as optimal
for spermatozoa [11±13], fertility was maintained at acceptable levels. However, possible
ways of reducing other eventual negative effects of a low semen/extender ratio would be to
increase the amount of known protective ingredients in the extender, such as egg yolk, or
the addition of a cryoprotectant at an earlier step during cooling, for instance before the
temperature reaches 5 8C. Several studies [14±16] have shown that cold shock starts soon
after the semen is cooled to room temperature, although it is more critical for ram semen at
temperatures below 15 8C [17]. Glycerol, despite its value as cryoprotectant, is metabo-
lically toxic to spermatozoa, and noxious to membrane integrity, depending on the
concentration and the temperature at which it is added, thus calling for a compromise
if optimal results are to be achieved [18±21].
Using animal-derived additives such as milk or egg yolk in a semen extender implies
sanitary risks, not only through the inclusion of speci®c microbiological agents, but also by
contaminants that may compromise the quality of the product [22,23]. This topic, frequently
discussed in the cattle AI industry, is of utmost importance in sheep AI as well. An
alternative to egg yolk in extenders for ram semen may be soybean lecithin, present in a
commercial extender developed for bull semen (Bioexcell1; IMV, L'Aigle, France).
 J. Gil et al. / Theriogenology 59 (2003) 1241±1255 1243

Therefore, the objective of the present work was to compare the effect of a milk-based
extender (Experiment 1), containing four different egg yolk concentrations (5, 10, 15, and
20%, respectively), with that of the commercial Bioexcell1 extender, containing two
different ®nal glycerol concentrations (3.2 and 6.4%), on post-thaw ram sperm character-
istics. In both experiments, the effect of adding glycerol at either 5 or 15 8C together with
the second fraction of each extender was also studied.

2. Materials and methods

2.1. Animals and semen collection

2.1.1. Experiment 1
In this experiment, semen from ®ve crossbred rams, 10±12 months of age, was used. The
rams were housed at the Department of Obstetrics and Gynecology, at the Swedish
University of Agricultural Sciences (SLU), Uppsala, for 2 months during the breeding
season (October to November). Semen was collected daily using an arti®cial vagina,
beginning 1 month before the start of the experiment. Two consecutive ejaculates were
collected within 10 min, and later pooled and treated as a single ejaculate. The ejaculates
were within normal ranges regarding volume (0.75±2 ml), concentration ( 2:5  109
spermatozoa/ml), motility (70%), and morphology (10% total sperm abnormalities).

2.1.2. Experiment 2
Semen from ®ve rams (®ve different breeds), 20±24 months of age, was used in this
experiment. The rams were housed at the Swedish Sheep Farmers Associations ram AI
station (KungsaÈngen, Uppsala) and involved in a daily semen collection routine during the
northern hemisphere breeding season (October to November). Only the second ejaculates,
collected over 5 consecutive days in November, were used. As in Experiment 1, all
ejaculates used were within the ranges of normality, as stated earlier.

2.2. Extenders

All extenders were prepared as follows, and stored at 18 8C until thawed and used in
the experiments.

2.2.1. Experiment 1
The effect of four milk-based extenders (M1, M2, M3, and M4) was studied. Because of
the two-step extension methodology used, each of the four extenders comprised two
fractions, the ®rst without glycerol and the second with 14% (v/v) glycerol, thus resulting
in a ®nal concentration of 7% (v/v) glycerol. Extenders were prepared from nonfatty milk
powder (11%, w/v) and distilled water. The milk was heated to 95 8C for 10 min on a
stirring plate equipped with a heater and then cooled to room temperature for further
processing. Fraction 1 was prepared by adding enough egg yolk to the milk-based
extender to reach concentrations of 5% (M5), 10% (M10), 15% (M15), and 20% (M20)
(v/v). After homogenization of the milk±egg yolk extender, all f/1 preparations were
1244 J. Gil et al. / Theriogenology 59 (2003) 1241±1255

centrifuged for 20 min at 3310  g at 5 8C to clarify the extender by removing large

particles and overlaying lipids. The clari®ed supernatant was supplemented with
antibiotics (penicillin 0.03 g/100 ml and streptomycin 0.04 g/100 ml). All f/2 were
prepared by adding 224 mM of fructose, egg yolk (5, 10, 15, 20%, v/v), and glycerol
(14%, v/v) to the milk base prepared, and supplemented with antibiotics, as described
earlier.

2.2.2. Experiment 2
Three extenders were used. The ®rst extender was a milk-based extender (M), prepared
as M5 in Experiment 1, and used as the control. The second extender (B) was the
Bioexcell1 extender (IMV, L'Aigle, France), which is commercially available for bull
semen. Due to the need for a two-step extension procedure, special preparations were
manufactured (IMV, L'Aigle, France), with f/1 without glycerol and f/2 with 6.4% (v/v)
glycerol. The third extender (BB) was the same as extender B, except that the f/2 contained
12.8% (v/v) glycerol.

2.3. Semen processing

After collection, the ejaculates were placed in a portable waterbath at 33 8C. An aliquot
from each ejaculate was taken to measure the sperm concentration in a photometer
(SpermaCue; MinituÈb, Tiefenbach, Germany). Each ejaculate from each ram was initially
split into as many fractions as needed for the experiments (Experiment 1 ˆ 4, Experiment
2 ˆ 3) (Fig. 1), and then extended 1 ‡ 1 with f/1 of extenders M5, M10, M15, and M20,
respectively (Experiment 1), and with f/1 of extenders M, B, and BB, respectively
(Experiment 2). After photometrically measuring the sperm concentrations, aliquots of
the preliminary extended semen, with equal sperm numbers from each ram, were then
pooled within each extender, and the pools were further extended to 1:6  109 cells/ml
with f/1 of each extender. After careful homogenization of the primary pools, the extended
semen was again split into two secondary pools, and further processed according to one of
the following protocols (see Fig. 1).
Protocol 5 8C (P5 8C):
1. Cooling to 5 8C within 60 min in a waterbath.
2. Dropwise extension at 5 8C with f/2 of each extender to 0:8  109 cells/ml.
3. Equilibration at 5 8C for 2 h.
4. Packaging in 0.25 ml mini-straws at 5 8C.
5. Freezing.

Protocol 15 8C (P15 8C):

1. Cooling to 15 8C within 30 min.

2. Dropwise extension at 15 8C with f/2 of each extender to 0:8  109 cells/ml.
3. Further cooling to 5 8C within 30 min in a waterbath.
4. Further equilibration at 5 8C for 1.5 h.
5. Packaging in 0.25 ml mini-straws at 5 8C.
6. Freezing.
 J. Gil et al. / Theriogenology 59 (2003) 1241±1255 1245

Fig. 1. Experimental design. In Experiment 1, the effect on sperm quality after thawing was studied with a
split-sample design (4  2 layout), using four extenders (M5, M10, M15, and M20) and addition of the
glycerolized fraction at either 5 or 15 8C. In Experiment 2, three extenders (M, B, and BB) were evaluated
using the same two protocols as in Experiment 1. The procedure shown in ®gure was repeated ®ve times in both
experiments.

For practical reasons, during transport from the ram AI station to the freezing laboratory,
the secondary pools were cooled to 15 8C within 30 min, either in a waterbath (Experiment 1)
at the freezing laboratory (Department of Obstetrics and Gynecology, SLU) or in a portable
isothermic chamber with wrapped freeze-packs (Experiment 2) to mimic the cooling rate
of the waterbath.
The freezing was done in a programmable freezing chamber (Digitcool 5300; IMV,
L'Aigle, France). Here, the temperature was lowered from 5 to 8 8C at a rate of 3 8C/min,
and from 8 to 130 8C at 25 8C/min. The straws were then transferred to LN2 for storage
until evaluated. In total, ®ve freezing operations for each experiment were done.

2.4. Experimental design

Two experiments were performed using a split-sample design. In Experiment 1 (4  2

layout), the effect on sperm quality post-thaw was studied using four extenders (M5, M10,
M15, and M20) with the glycerolized fraction being added at 5 8C (P5 8C) or at 15 8C
(P15 8C), respectively (Fig. 1). In Experiment 2 (3  2 layout), three extenders (M, B,
and BB) were studied using the same two protocols for the addition of the glycerolized
fraction of each extender.
1246 J. Gil et al. / Theriogenology 59 (2003) 1241±1255

2.5. Post-thaw semen evaluation

2.5.1. Thawing of the straws

The straws were thawed at 50 8C for 9 s in a waterbath. For each evaluation, three straws
were thawed and pooled in a test tube.

2.5.2. Subjective motility (SM)

This was assessed post-thaw using a phase contrast microscope equipped with a warm
stage (38 8C) (Nikon, Tokyo, Japan). Ten microliters of pooled semen was diluted at 38 8C
with 100 ml of sucrose buffer (10 mM NaCl, 222 mM sucrose, 2.5 mM KCl, 20 mM
HEPES, 10 mM glucose, and 1 mg/100 ml of cold water-soluble polyvinyl alcohol (pH 7.5
with NaOH 1N; Sigma Aldrich, TyresoÈ, Sweden)). Then, 5 ml of the diluted sample was
placed in a Makler Counting Chamber (Se®-Medical Instruments, Haifa, Israel), and the
frequency of sperm motility assessed to the nearest 5%, after judging four different
microscopic ®elds.

2.5.3. Cell motion analysis (CASA)

For the same preparation of diluted semen and with the same microscope as used for SM
(see earlier description), CASA was performed with a computer-assisted motility analysis
system (StroÈmberg-Mika Cell Motion Analyzer; SM-CMA, MTM Medical Technologies,
Montreux, Switzerland). For each evaluation, eight microscopic ®elds (sequences) were
analyzed to include at least 200 spermatozoa. The proportion of total motile (TM) and
linearly motile (LM) spermatozoa, straight-line velocity (VSL, mm/s), curvilinear velocity
(VCL, mm/s), average-path velocity (VAP, mm/s), and spermatozoa with lateral head
displacement (LHD, mm) was determined. The main software settings were 32 frames per
sequence, minimum of 15 frames per object, 10 mm/s as a velocity limit for immobile
objects, 25 mm/s as a velocity limit for local motile objects, and 25 mm for the maximum
radius of circles.

2.5.4. Sperm membrane integrity

This was assessed using staining with SYBR-14 and propidium iodide (SYBR-14/PI;
Molecular Probes Inc., Eugene, OR, USA), as described by Garner and Johnson [24].
Thawed semen (50 ml) was mixed with 5 ml of SYBR-14 and 2.5 ml of PI (10 mM SYBR-14
and 120 mM PI). The sample was then gently mixed and incubated at 37 8C for 15 min. The
smears were prepared by placing 4 ml of stained semen and 1 ml of glutaraldehyde solution
(3% electron microscopy grade in cacodylate buffer, pH 7.4) on a warm slide just before the
assessment in order to stop the motion, thus facilitating counting. Two hundred spermatozoa
were counted using a microscope (Leitz Laborlux-11; Jena, Germany) equipped with a
Paralens1 objective set (60, 470±490 nm excitation ®lter, 510 nm dichroic beam splitter,
520 nm barrier ®lter; Becton-Dickinson, Leiden, The Netherlands) and classi®ed as `intact'
when stained green or as `membrane damaged' when stained green±red or just red.

2.5.5. Capacitation status

This was assessed in viable spermatozoa with chlortetracycline (CTC), as described by
Gil et al. [9,10,25]. To assess only the viable spermatozoa, 18 ml of thawed spermatozoa
 J. Gil et al. / Theriogenology 59 (2003) 1241±1255 1247

was incubated for 10 min at 37 8C with 2 ml of ethidium homodimer (23.3 mM EthD-1,

Molecular Probes Inc., Eugene, OR, USA). After incubation, the mixture was ®xed with
20 ml of 3% glutaraldehyde (electron microscopy grade in cacodylate buffer, pH 7.4), and
then stained with 40 ml of CTC working solution. The sample was washed by centrifuga-
tion at 700  g for 8 min with 1 ml of sucrose buffer (see earlier description). Thereafter,
1 ml of the supernatant was removed and the pellet re-suspended. The smears were
prepared with 4 ml of stained sample plus 2 ml of antifade (0.1% p-pheneylendiamine in
9 ‡ 1 of glycerol and phosphate-buffered saline (PBS)), mixed on a clean microscope
slide, covered with a cover glass, sealed with nail varnish, and kept in the dark at 4 8C. The
evaluations were done, within 12 h of staining, using a microscope (Leitz Diaplan-20, Jena,
Germany) with epi¯uorescent optics and a set of violet±blue ®lters (420±490 nm excita-
tion, 510 nm emission) and green ®lters (530±560 nm excitation, 580 nm emission). Two
hundred viable spermatozoa (unstained with EthD-1) were classi®ed into three categories:
the uniform ¯uorescent head (uncapacitated: CTC-F), the ¯uorescence-free band in the
post-acrosomal region (capacitated: CTC-B), and the non¯uorescent head or a thin
¯uorescent band in the equatorial segment (acrosome-reacted: CTC-AR) [9,26].

2.6. Statistical analysis

The results of each experiment were analyzed using the General Linear Model (GLM) in
the Statistical Analysis System package (SAS Institute, Inc., Cary, NC, USA, 1996). The
results are presented as least square means …LSM†  standard error of the means (S.E.M.).
In the statistical model, the classes de®ned were the freezing operation, extender, freezing
protocol, and extender  freezing protocol interaction. Due to the signi®cant differences
found for the freezing operation in Experiment 1, a further mixed procedure was used
including the effect of the freezing operation as a random effect.

3. Results

3.1. The effect of extenders on sperm parameters

3.1.1. Experiment 1
For motility parameters, the SM values (%) obtained were higher (although NS) for M5
and M10 than for M15 and M20 (70  2:2, 70  2:2, 68  2:2, and 67  2:2, respectively).
The TM values (%) were almost equal for all extenders (83  2:3, 83  2:3, 84  2:3, and
83  2:3), as were the frequencies of LM spermatozoa assessed by CASA (M5: 65  2:8%,
M10: 62  2:8%, M15: 60  2:8%, and M20: 59  2:8%). The velocity patterns (VSL,
VCL, VAP, and LHD) assessed did not differ signi®cantly between extenders.
The percentages of spermatozoa with intact membranes (SYBR-14/PI) were lower in
M5 and M20 than in M10 and M15 (M5: 64  2:1, M10: 69  2:1, M15: 66  2:1, M20:
63  2:1). The overall effect of the extenders on membrane intactness (P < 0:01) was
affected by the extender used, and the comparison between extenders showed statistically
signi®cant differences between M5 and M10 and between M10 and M20 (P < 0:05 to
P < 0:01).
1248 J. Gil et al. / Theriogenology 59 (2003) 1241±1255

The percentage of uncapacitated spermatozoa assessed by CTC (CTC-F) was signi®-

cantly affected (P < 0:01) by the extender used. The highest value was seen in M5
(51  2:4%) compared with the other extenders (M10: 41  2:4%, M15: 39  2:4%, and
M20: 41  2:4%). Conversely, the population of capacitated spermatozoa (CTC-B) was
lower in M5 extender (44  2:6%; P < 0:01) than in all the others (M10: 51  2:6%, M15:
53  2:6%, and M20: 52  2:6%). The percentage of acrosome-reacted spermatozoa
(CTC-AR) did not differ between extenders.

3.1.2. Experiment 2
Overall, the motility (SM, TM, and LM) was signi®cantly (P < 0:001) in¯uenced by
extenders. The pairwise test of SM showed signi®cant differences between extenders B
(26  2:9%) and M and BB, but not between M and BB (58  2:9 and 53  2:9%,
respectively). The percentage of TM spermatozoa was signi®cantly (P < 0:001) lower in B
(28  4:2%) than in M (67  4:2%) and BB (63  4:2%). The percentages of LM
spermatozoa were signi®cantly (P < 0:001) lower for extender B (21  3:3%) than for
BB (45  3:3%) and M (44  3:3%). The VSL values differed signi®cantly (P < 0:001)
between extenders BB and B as well as between BB and M, but not between B and M, with
a higher value for B (115  3:8) than for M and BB (114  3:8 and 102  3:8, respec-
tively). The VAP (mm/s) for extender M was higher (126  4:1) than for B (123  4:1) and
BB (111  4:1), and the difference was signi®cant (P < 0:01) between BB and the other
two extenders. The VCL (mm/s) was higher for spermatozoa frozen in M (159  4:1) than
for those frozen in B (141  4:1) and BB (135  4:1), and the difference was signi®cant
(P < 0:01) between M and the other two extenders.
Overall, the percentage of spermatozoa with intact membranes was in¯uenced by
extenders (P < 0:001). The proportion was higher in samples frozen in extender M
(57  3:1%) than in those frozen in extenders BB (46  3:1%) and B (33  3:1%).
The pairwise comparison showed signi®cant differences (P < 0:001) between the three
extenders. The proportion of CTC-F spermatozoa (uncapacitated) were also affected by
extenders (P < 0:001), where extender M (44  2:4%) showed signi®cantly higher
(P < 0:001) values than did BB (30  2:4%) and B (30  2:4%). No signi®cant difference
was found between B and BB. Conversely, the proportion of CTC-B (i.e. capacitated
spermatozoa) was lower (P < 0:001) in extender M (48  1:8%) than in BB (50  1:8%)
and in B (58  1:8%). No signi®cant difference was found between M and BB. The
proportion of CTC-AR spermatozoa was lower in extender M (8  1:4) than in B
(11  1:4) and in BB (19  1:4), and signi®cant differences were found between all
extenders (P < 0:05 to P < 0:001).

3.2. The effect of the freezing protocol on sperm parameters

3.2.1. Experiment 1
Addition of the f/2 (glycerolized fraction) at 15 8C (69  2:1%) did not affect the SM
compared to addition at 5 8C (69  2:1%). The proportion of TM spermatozoa assessed by
CASA did not differ signi®cantly between P5 8C and P15 8C (84  1:8% versus
82  1:8%), nor did the proportion of LM spermatozoa (60  2:4% versus 62  2:4%).
No signi®cant difference in velocity patterns was found between protocols.
 J. Gil et al. / Theriogenology 59 (2003) 1241±1255 1249

The percentage of spermatozoa having intact membranes was signi®cantly higher

(P < 0:01) in the samples processed using the P5 8C protocol than in those processed
using P15 8C (68  1:8% versus 63  1:8%). Regarding the capacitation status, the CTC-F
value was slightly higher (NS) in P5 8C than in P15 8C (44  1:9% versus 42  1:9%). The
CTC-B values were equal in both protocols (50  2:3%) and the CTC-AR spermatozoa
values were signi®cantly lower (P < 0:01) in P5 8C than in P15 8C (6  0:9% versus
8  0:9%).

3.2.2. Experiment 2
The addition of f/2 (with glycerol) at 5 8C yielded higher (P < 0:05) percentages of SM
compared to addition at 15 8C (48  2:6% versus 42  2:6%). Furthermore, the percentage
of TM spermatozoa was higher (NS) in P5 8C than in P15 8C (56  3:6% versus
51  3:6%), and the percentage of LM spermatozoa was higher (although NS) in
P5 8C than in P15 8C (39  2:9% versus 35  2:9%). The velocity patterns VSL, VAP,
and VCL did not differ between the two protocols for addition of glycerol.
The percentage of membrane-intact spermatozoa was higher (NS) for P5 8C than for
P15 8C (47  2:9% versus 44  2:9%). Finally, the percentage of CTC-F spermatozoa was
slightly higher (NS) for P5 8C (36  2:2%) than for P15 8C (35  2:2%), the values for
CTC-B spermatozoa were equal (52  1:6%), and the proportion of CTC-AR spermatozoa
was higher (NS) in P15 8C than in P5 8C (13  1:2% versus 12  1:2%).

3.3. The effect of the interaction between freezing protocol and extender
on the sperm parameters

3.3.1. Experiment 1
The extender  protocol interaction did not in¯uence either the motility (SM, LM, and
TM; Table 1) or the sperm velocity parameters assessed by CASA (VSL, VAP, and VCL;
Table 2).

Table 1
Motility parameters in frozen±thawed ram spermatozoa (max/min (LSM)) for two freezing protocols (P5 8C
and P15 8Ca) and four milk extenders (M5, M10, M15, and M20b)

Extender Protocol Subjective motile Total motile Linear motile

(S.E.M. 2.5) (CASA; S.E.M. 3.1) (CASA; S.E.M. 3.6)

M5 P5 8C 74/68 (71) 89/76 (84) 69/46 (63)

P15 8C 75/65 (70) 92/77 (82) 78/59 (66)
M10 P5 8C 76/65 (71) 90/78 (85) 67/52 (61)
P15 8C 76/66 (70) 88/70 (81) 72/56 (62)
M15 P5 8C 78/61 (69) 94/78 (84) 75/50 (60)
P15 8C 74/58 (68) 89/75 (84) 68/52 (61)
M20 P5 8C 74/58 (66) 89/69 (83) 65/49 (57)
P15 8C 74/58 (69) 91/67 (82) 74/48 (61)
a
With the addition of the glycerolized fraction at 5 and 15 8C, respectively.
b
Containing 5, 10, 15, and 20% egg yolk, respectively.
1250 J. Gil et al. / Theriogenology 59 (2003) 1241±1255

Table 2
Velocity (mm/s) in frozen±thawed ram spermatozoa assessed by CASA (max/min (LSM)) for two freezing
protocols (P5 8C and P15 8Ca) and four milk extenders (M5, M10, M15, and M20b)

Extender Protocol Straight-line Average-path Circular-line

(VSL; S.E.M. 4.1) (VAP; S.E.M. 3.9) (VCL; S.E.M. 3.8)

M5 P5 8C 150/120 (138) 161/136 (151) 193/174 (184)

P15 8C 158/126 (142) 170/134 (151) 201/163 (180)
M10 P5 8C 142/120 (133) 153/131 (144) 186/167 (179)
P15 8C 151/123 (138) 163/133 (148) 200/167 (181)
M15 P5 8C 148/124 (133) 155/139 (145) 190/174 (179)
P15 8C 147/124 (138) 158/140 (150) 191/176 (184)
M20 P5 8C 140/112 (129) 155/125 (142) 192/158 (177)
P15 8C 152/116 (135) 161/129 (146) 195/160 (178)
a
With the addition of the glycerolized fraction at 5 and 15 8C, respectively.
b
Containing 5, 10, 15, and 20% egg yolk, respectively.

The overall extender  protocol interaction for the extenders signi®cantly in¯uenced the
membrane integrity (P < 0:01), and the pairwise comparison within extender showed
higher proportions of spermatozoa with intact membranes for all extenders processed using
the P5 8C protocol than for P15 8C, but the difference was only signi®cant for M15
(70  2:6% versus 63  2:6%; P < 0:05). Regarding the capacitation status, no difference
in the CTC-F or in the CTC-B was found. There was only a tendency for a difference
(P ˆ 0:06) between protocols within M15 for CTC-B (P5 8C: 57  3:2% versus P15 8C:
50  3:2%). The overall CTC-AR was higher (P < 0:01) with P15 8C than with P5 8C
(6  1:0% versus 8  1:0%), mainly due to the difference found in extender M10 when

Table 3
Percentages (max/min (LSM)) of frozen±thawed ram spermatozoa showing sperm membrane integrity (SYBR-
14/PI) and phases of capacitation (CTC) for two freezing protocols (P5 8C and P15 8Ca) and four milk extenders
(M5, M10, M15, and M20b)

Extender Protocol Membrane integrity Uncapacitated Capacitated Acrosome-reacted

(S.E.M. 2.6) (S.E.M. 3.1) (S.E.M. 3.2) (S.E.M. 1.4)

M5 P5 8C 69/61 (65) 60/47 (53) 50/35 (42) 8/3 (5)

P15 8C 71/54 (63) 55/49 (48) 56/40 (47) 7/3 (5)
M10 P5 8C 75/65 (71) 48/39 (44) 58/40 (51) 12/3 (5)
P15 8C 72/61 (67) 52/26 (38) 61/38 (51) 13/7 (10)**
M15 P5 8C 82/61 (70)* 48/30 (39) 68/49 (57) 9/2 (4)
P15 8C 71/57 (63) 53/32 (40) 62/37 (50) 17/4 (10)**
M20 P5 8C 70/61 (65) 44/37 (41) 58/45 (52) 12/4 (7)
P15 8C 70/49 (61) 48/30 (42) 61/48 (52) 9/3 (6)
a
With the addition of the glycerolized fraction at 5 and 15 8C, respectively.
b
Containing 5, 10, 15, and 20% egg yolk, respectively.
*
Denotes signi®cant difference (P < 0:05) within extender between protocols.
**
Denotes signi®cant difference (P < 0:01) within extender between protocols.
 J. Gil et al. / Theriogenology 59 (2003) 1241±1255 1251

Table 4
Motility parameters in frozen±thawed ram spermatozoa (max/min (LSM)) for two freezing protocols (P5 8C and
P15 8Ca) and three extenders (M, B, and BBb)

Extender Protocol Subjective motile Total motile Linear motile

(S.E.M. 3.6) (CASA; S.E.M. 5.6) (CASA; S.E.M. 4.3)

M P5 8C 65/54 (58) 83/56 (65) 49/36 (43)

P15 8C 65/53 (57) 79/59 (69) 53/41 (46)
B P5 8C 46/24 (32)* 46/16 (33) 32/12 (24)
P15 8C 40/9 (20) 51/12 (24) 43/8 (19)
BB P5 8C 64/45 (56) 83/49 (68) 65/31 (51)*
P15 8C 58/39 (50) 76/41 (59) 50/28 (39)
a
With the addition of the glycerolized fraction at 5 and 15 8C, respectively.
b
M: milk, B: Bioexcell1 3.2% glycerol, and BB: Bioexcell1 6.4% glycerol.
*
Denotes a signi®cant difference (P < 0:05) within extender between protocols.

processed using the P15 8C as compared with P5 8C (10  1:4% versus 5  1:4%;
P < 0:01), and in extender M15 when processed using P15 8C compared with P5 8C
(10  1:4% versus 4  1:4%; P < 0:01). However, no difference between protocols was
seen for extenders M5 or M20 (Table 3).

3.3.2. Experiment 2
The extender  protocol interaction in¯uenced sperm motility when using extenders B
and BB, but to a lesser extent in samples processed in extender M (Table 4). Within the
extenders generally, the SM was higher (P < 0:05) in samples processed using the P5 8C
protocol than in those using P15 8C (48  2:6% versus 42  2:6%). A pairwise comparison
showed signi®cant differences (P < 0:05) between protocols only in B (P5 8C: 32  3:6%
versus P15 8C: 20  3:6%). The TM values did not differ signi®cantly between protocols.
The interaction did not in¯uence the proportions of LM spermatozoa, except within BB
(P5 8C: 51  4:3% versus P15 8C: 39  4:3%; P < 0:05). No signi®cant differences in the
velocity patterns VSL, VAP, and VCL were found due to the interaction (Table 5).

Table 5
Velocity (mm/s) in frozen±thawed ram spermatozoa assessed by CASA (max/min (LSM)) for two freezing
protocols (P5 8C and P15 8Ca) and three extenders (M, B, and BBb)

Extender Protocol Straight-line (VSL) Average-path Circular-line

(S.E.M. 4.5) (VAP; S.E.M. 4.8) (VCL; S.E.M. 5.0)

M P5 8C 134/146 (115) 146/114 (128) 186/152 (165)

P15 8C 118/99 (113) 129/111 (124) 167/138 (154)
B P5 8C 123/92 (111) 133/98 (119) 152/116 (137)
P15 8C 133/114 (120) 138/119 (126) 159/137 (145)
BB P5 8C 115/89 (104) 126/98 (113) 150/123 (136)
P15 8C 112/90 (100) 124/100 (110) 147/122 (133)
a
With the addition of the glycerolized fraction at 5 and 15 8C, respectively.
b
M: milk, B: Bioexcell1 3.2% glycerol, and BB: Bioexcell1 6.4% glycerol.
1252 J. Gil et al. / Theriogenology 59 (2003) 1241±1255

Table 6
Percentages (max/min (LSM)) of frozen±thawed ram spermatozoa depicting sperm membrane integrity (SYBR-
14/PI) and phases of capacitation (CTC) for two freezing protocols (P5 8C and P15 8Ca) and three extenders (M,
B, and BBb)

Extender Protocol Membrane integrity Uncapacitated Capacitated Acrosome-reacted

(S.E.M. 3.5) (S.E.M. 2.8) (S.E.M. 2.3) (S.E.M. 1.7)

M P5 8C 61/46 (57) 53/35 (44) 56/40 (49) 13/4 (8)

P15 8C 65/47 (58) 51/40 (45) 52/44 (47) 11/5 (8)
B P5 8C 39/28 (34) 40/24 (33) 62/48 (57) 14/6 (10)
P15 8C 44/27 (33) 35/22 (29) 68/55 (59) 17/11 (12)
BB P5 8C 58/40 (50) 35/26 (31) 55/47 (51) 28/13 (19)
P15 8C 58/28 (42) 39/18 (31) 59/44 (50) 24/16 (20)
a
With the addition of the glycerolized fraction at 5 and 15 8C, respectively.
b
M: milk, B: Bioexcell1 3.2% glycerol, and BB: Bioexcell1 6.4% glycerol.

The interaction did not in¯uence the overall values for spermatozoa with an intact
membrane, but the percentages of spermatozoa with an intact membrane differed sig-
ni®cantly (P < 0:05) within extender BB (P5 8C: 50  3:5% versus P15 8C: 42  3:5%).
None of the categories used for assessing the capacitation status was affected by the
interaction between protocol and extender (Table 6).

4. Discussion

The results of the present study show that increasing egg yolk concentration higher than
5% in milk extender does not signi®cantly improve sperm characteristics after thawing,
except for the percentage of membrane intactness, where the preparation containing 10%
egg yolk gave signi®cantly better results. Our results also indicate that Bioexcell1 with a
glycerol concentration of 6.4% glycerol yielded similar results as did the control extender
used (milk-based, with 5% egg yolk).
The addition of f/2 at 15 8C did not signi®cantly improve the sperm quality post-thaw.
Conversely, membrane intactness was signi®cantly higher at 5 8C compared to 15 8C and
the proportion of acrosome-reacted spermatozoa was lower at 5 8C when using milk-based
extenders. Although the differences were not always statistically signi®cant for the sperm
characteristics studied, all parameters were better when adding f/2 with glycerol at 5 8C,
regardless of the egg yolk (in milk extenders) or glycerol (in Bioexcell1) concentration
used. Similar results were obtained in previous studies [19].
Mortimer and Maxwell [27] de®ned the kinematics of hyperactivated ram spermatozoa
from a subjectively selected sperm population by depicting a hyperactivation pattern with a
decreased linearity and VSL, plus an increased VCL and the amplitude of the LHD as the
main components. Furthermore, sperm hyperactivation is a phenomenon associated with
capacitation [28]. In the present study, we found that the percentages of capacitated
spermatozoa increased with an increase in the egg yolk concentration in extender M as well
as in Bioexcell1 containing 3.2% rather than 6.4% glycerol, concomitantly with decreased
 J. Gil et al. / Theriogenology 59 (2003) 1241±1255 1253

frequencies for the motility patterns (VSL, LM, and SM) evaluated. In our study, the CASA
analysis took into account all detected spermatozoa without any kind of selection as in the
study by Cummings [28]. Consequently, we did not ®nd any differences in the velocity and
motility patterns, but all the values decreased progressively, in absolute ®gures, as the
concentration of egg yolk in the extenders was increased from 5 to 20%. The protective
effect of egg yolk on sperm membrane integrity is largely recognized in farm animals, but it
may also be involved in membrane destabilization, as shown by an increased frequency of
acrosome damage with an increase in the concentration of egg yolk [29±31]. Despite this,
egg yolk has been ruled out as a capacitating factor in itself [32]. This may partly explain
why, in the present study, extender M10 had fewer uncapacitated spermatozoa than did M5,
despite the higher percentage of spermatozoa with intact membranes. Capacitation
represents a destabilization of the sperm membrane, which submits more readily to
spontaneous acrosome exocytosis, thus lowering the half-life of the cell population
[16]. This side effect of egg yolk may explain the decreased proportion of ram spermatozoa
with intact membranes in the extender with 20% egg yolk found in the present study.
Sperm membrane parameters, such as intactness and stability, were better in P5 8C than
in P15 8C. This could be explained by the reported effect of glycerol as a promoter of
capacitation-like changes [33], or by its toxic effect on sperm metabolism [21]. The
frequencies of stable (uncapacitated) and capacitated spermatozoa were almost the same in
P5 8C and P15 8C; however, the subpopulations showing acrosome reactions were sig-
ni®cantly higher in P15 8C than in P5 8C. All tested sperm attributes seemed to be better
when adding f/2 at 5 8C than at 15 8C, except in Experiment 2 for the control (M5) and
extender BB, where there was almost no difference between the protocols.
Concerns about using egg yolk in semen extenders have been a major issue in
discussions in the cattle AI industry in recent years and some studies with relevant
substitutes have been conducted to address the effect of extenders [22,25,34,35]. Similar
concerns also apply to sheep AI. Apart from sanitary concerns, there is a possibility that
milk and egg yolk could alter the chromatin structure of the spermatozoa, and thus reduce
sperm quality post-thaw [36]. In this study, we have followed the concept of ``wherever
possible, additives of animal origin are to be avoided'' [23], by testing the commercially
available extender Bioexcell1, with two different glycerol concentrations (®nal concen-
tration: 3.2 and 6.4%, respectively). Most of the variables yielded better results in
Bioexcell1 with a concentration of 6.4% of glycerol (BB) than in Bioexcell1 containing
3.2% glycerol. The latter extender (B) also showed lower values for the sperm quality
parameters assessed than did the control (milk extender with 5% egg yolk). According to
the kinematics of spermatozoa showing hyperactivation, and the results of the CTC test, the
spermatozoa processed in extender B underwent a more stressful treatment than did those
in BB, probably due to a too low concentration of glycerol. These ®ndings are also
underlined by the lower percentage of membrane integrity found in extender B. The fact
that BB achieved similar, or even better, values (considering LM) than did the control (milk
extender) indicates that this extender, free from additives of animal origin, may be used as
an alternative extender when freezing ram semen.
In conclusion, an increase of the egg yolk content above 10% in the milk-based extender
did not represent any improvement in the post-thaw sperm quality. Bioexcell1 with a ®nal
glycerol concentration of 6.4% may be a good alternative, free from additives of animal
1254 J. Gil et al. / Theriogenology 59 (2003) 1241±1255

origin, to the traditional milk extender (supplemented with 5% egg yolk) for freezing ram
semen. The addition of glycerol at 15 8C compared to that at 5 8C did not present any
improvement in the sperm quality after thawing. However, one must remember that the
results presented here are based on in vitro evaluations. Further studies must be performed
in order to evaluate whether the extenders and freezing conditions tested herein have any
in¯uence on the fertility results after cervical AI under ®eld conditions.

Acknowledgements

We would like to thank Dr. Decuadro Hansen of IMV at L'Aigle, France, for the
preparation of Bioexcell1. The authors received ®nancial support for this work from the
Swedish Agency for Research Cooperation with Developing Countries (SAREC) and
KoÈttboÈndernas Forskningsprogram, Stockholm, Sweden. Dr. Jorge Gil is on leave of
absence from the Veterinary Laboratories ``M.C. Rubino'' (DILAVE), Ministry of Live-
stock, Agriculture and Fisheries (MGAP), Uruguay.

References

[1] Salamon S, Maxwell WMC. Frozen storage of ram semen. I. Processing, freezing, thawing and fertility
after cervical insemination. Anim Reprod Sci 1995;37:185±249.
[2] Evans G, Maxwell WMC. Current status of embryo transfer and arti®cial insemination in small ruminants.
In: Proceedings of the Satellite Symposium on Reproduction in Small Ruminants, 14th ICAR Congress,
Session 3. Sandnes: Norway; 2000, p. 54±9.
[3] El-Alamy MA, Foote RH. Freezability of spermatozoa from Finn and Dorset rams in multiple semen
extenders. Anim Reprod Sci 2001;65:245±54.
[4] Ollero M, Perez-Pe R, Muino-Blanco T, Cebrian-Perez A. Improvement of ram sperm cryopreservation
assessed by sperm quality parameters and heterogeneity analysis. Cryobiology 1998;37:1±12.
[5] Holt WV. Basic aspects of frozen storage of semen. Anim Reprod Sci 2000;62:3±22.
[6] Holt WV. Fundamental aspects of sperm cryobiology: the importance of species and individual
differences. Theriogenology 2000;53:47±58.
[7] Curry MR, Kleinhans FW, Watson PF. Measurement of the water permeability of the membranes of boar,
ram, and rabbit spermatozoa using concentration-dependent self-quenching of an entrapped ¯uorophore.
Cryobiology 2000;41:167±73.
[8] SoÈderquist L, Madrid-Bury N, Rodriguez-Martinez H. Assessment of membrane integrity after using
different procedures to thaw ram spermatozoa frozen in mini-straws. Theriogenology 1997;48:1115±25.
[9] Gil J, SoÈderquist L, Rodriguez-Martinez H. In¯uence of centrifugation and different extenders on post-
thaw sperm quality of ram semen. Theriogenology 2000;54:93±108.
[10] Gil J, Rodriguez-Irazoqui M, SoÈderquist L, Rodriguez-Martinez H. In¯uence of centrifugation or low
prefreezing extension rates on the fertility of ram semen after cervical insemination. Theriogenology
2002;57:1781±92.
[11] Martin IC, Watson PF. The effects of dilution rate and diluent on the fertility of ram semen. J Reprod Fertil
1973;32:310±1.
[12] Salamon S, Maxwell WMC. Storage of ram semen. Anim Reprod Sci 2000;62:77±111.
[13] D'Alessandro AG, Martemucci AG, Colonna MA, Bellitti A. Post-thaw survival of ram spermatozoa and
fertility after insemination as affected by prefreezing sperm concentration and extender composition.
Theriogenology 2001;55:1159±70.
[14] Salamon S, Maxwell WMC. Frozen storage of ram semen. II. Causes of low fertility after cervical
insemination and methods of improvement. Anim Reprod Sci 1995;38:1±36.
 J. Gil et al. / Theriogenology 59 (2003) 1241±1255 1255

[15] Maxwell WM, Watson PF. Recent progress in the preservation of ram semen. Anim Reprod Sci 1996;42:
55±65.
[16] Watson PF. The causes of reduced fertility with cryopreserved semen. Anim Rep Sci 2000;60:481±92.
[17] Fiser PS, Fairfull RW. The effects of rapid cooling (cold shock) of ram semen, photoperiod, and egg yolk
in diluents on the survival of spermatozoa before and after freezing. Cryobiology 1986;23:518±24.
[18] Colas G. Effect of initial freezing temperature, addition of glycerol and dilution on the survival and
fertilizing ability of deep-frozen ram semen. J Reprod Fertil 1975;42:277±85.
[19] Fiser PS, Fairfull RW. Combined effects of glycerol concentration, cooling velocity, and osmolality of
skim milk diluents on cryopreservation of ram spermatozoa. Theriogenology 1986;25:473±84.
[20] Fiser PS, Fairfull RW. The effects of glycerol-related changes on post-thaw motility and acrosomal
integrity of ram spermatozoa. Cryobiology 1989;26:64±9.
[21] Fahy GM. The relevance of cryoprotectant toxicity to cryobiology. Cryobiology 1986;23:1±13.
[22] Bousseau S, Brillard JP, Marquant-Le Guienne B, Guerin B, Camus A, Lechat M. Comparison of
bacteriological qualities of various egg yolk sources and the in vitro and in vivo fertilizing potential of
bovine semen frozen in egg yolk or lecithin based diluents. Theriogenology 1998;50:699±706.
[23] Thibier M, Guerin B. Hygienic aspects of storage and use of semen for arti®cial insemination. Anim
Reprod Sci 2000;62:233±51.
[24] Garner D, Johnson LA. Viability assessment of mammalian sperm using SYBR-14 and propidium iodide.
Biol Reprod 1995;53:276±84.
[25] Gil J, Januskauskas A, HaÊaÊrd MC, HaÊaÊrd MGM, Johannisson A, SoÈderquist L, et al. Functional sperm
parameters and fertility of bull semen extended in Biociphos Plus1 and Triladyl1. Reprod Domest Anim
2000;35:69±77.
[26] Gillan L, Evans G, Maxwell WM. Capacitation status and fertility of fresh and frozen-thawed ram
spermatozoa. Reprod Fertil Dev 1997;9:481±7.
[27] Mortimer ST, Maxwell WMC. Kinematic de®nition of ram sperm hyperactivation. Reprod Fertil Dev
1999;11:25±30.
[28] Cummings JM. Hyperactivated motility patterns of ram spermatozoa recovered from the oviducts of
mated ewes. Gamete Res 1982;6:53±63.
[29] Watson PF, Martin IC. The response of ram spermatozoa to preparations of egg yolk in semen diluents
during storage at 5 or 196 8C. Aust J Biol Sci 1973;26:927±35.
[30] Watson PF, Martin IC. Arti®cial insemination of sheep: the effect of semen diluents containing egg yolk
on the fertility of ram semen. Theriogenology 1976;6:559±64.
[31] Smith RL, Berndtson WE, Unal MB, Pickett BW. In¯uence of percent egg yolk during cooling and
freezing on survival of bovine spermatozoa. J Dairy Sci 1979;62:1297±303.
[32] Ijaz A, Hunter AG, Graham EF. Identi®cation of the capacitating agent for bovine sperm in egg yolk-
TEST semen extender. J Dairy Sci 1989;72:2700±6.
[33] Slavik T. Effect of glycerol on the penetrating ability of fresh ram spermatozoa with zona-free hamster
eggs. J Reprod Fertil 1987;79:99±103.
[34] Hinsch E, Hinsch KD, Boehm JG, Schill WB, Mueller-Schloesser F. Functional parameters and
fertilisation success of bovine semen cryopreserved in egg-yolk free and egg-yolk containing extenders.
Reprod Domest Anim 1997;32:143±9.
[35] Van Wagtendonk-de Leew AM, Haring RM, Kaal-Lansbergen LMTE, Den Daas JHG. Fertility results
using bovine semen cryopreserved with extenders based on egg yolk and soy bean extract. Theriogenology
2000;54:57±67.
[36] Karabinus DS, Evenson DP, Kaproth MT. Effects of egg yolk-citrate and milk extenders on chromatin
structure and viability of cryopreserved bull sperm. J Dairy Sci 1991;74:3836±48.