Sperm concentration at freezing affects post-thaw quality and fertility of ram
M. Alvarez1,2, J. Tamayo-Canul1,2, E. Anel1,2, J.C. Boixo4, M. Mata-Campuzano1,2, F.
Martinez-Pastor1,3, L. Anel 1,2 and P de Paz 1,3.
1ITRA-ULE, INDEGSAL, University of León, 24071, León, Spain
2Animal Reproduction and Obstetrics, University of León, 24071, León, Spain
3Molecular Biology, University of León, 24071, León, Spain.
4CENSYRA, León, Spain.
Campus de Vegazana
24071 León, Spain
Phone: +34 987291000+5679
We have investigated the effect of sperm concentration in the freezing doses (200, 400,
800 and 1600×106 mL-1) on the post-thaw quality and fertility of ram semen. Semen
was collected from seven adult Churra rams by artificial vagina during the breeding
season. The semen was diluted in a extender (TES-Tris-fructose, 20% egg yolk and 4%
glycerol), to a final concentration of 200, 400, 800 or 1600×106 mL-1 and frozen. Doses
were analyzed post-thawing for motility (CASA), viability and acrosomal status
(fluorescence probes PI/PNA-FITC, SYBR-14/PI and YO-PRO-1/PI). Total motility and
velocity were lower for 1600×106 mL-1 doses, while progressive motility and viability
were lower both for 800 and 1600×106 mL-1. The proportion of viable spermatozoa
showing increased membrane permeability (YO-PRO-1+) rised in 800 and 1200×106
mL-1. Intrauterine inseminations were performed with the 200, 400 and 800×106 mL-1
doses at a fixed sperm number (25×106 per uterine horn) in synchronized ewes. Fertility
(lambing rate) was similar for semen frozen at 200 (57.5%) or 400x106 mL-1 (54.4%),
whereas it was significantly lower for 800×106 mL-1 (45.5%). In conclusion, increasing
sperm concentration in cryopreserved semen, at least at 800×106 mL-1 and above,
adversely affects the post-thawing quality and fertility of ram semen.
Ram, sperm cryopreservation, sperm concentration, sperm quality, fertility
The efficiency of the cryopreservation of ram semen must be improved before
widespread application of artificial insemination (AI) in sheep. Acceptable results have
been achieved so far using frozen/thawed semen [1?4], but its general use is restricted
due to the need of using intrauterine insemination by laparoscopy. Otherwise, AI with
frozen semen yields variable and often low fertility results, if applied by vaginal-
cervical insemination [4?10]. Another disadvantage of vaginal AI is the high number of
spermatozoa required per insemination (100-400×106 spermatozoa/dose), whereas
laparoscopic AI requires lower sperm numbers (25-50×106 spermatozoa/dose) [4,11,12].
In fact, the effect of sperm dose in the cryopreservation of ram semen has been little
explored. To our knowledge, the only study was performed by D'Alessandro et al. ,
who tested two types of diluents (milk-lactose-egg yolk and Tris-fructose-egg yolk),
freezing at six different sperm concentrations (50, 100, 200, 400, 500 and 800×106 mL-
1). They found a variable sperm quality among 50 and 500×106 mL-1, but freezing at
800×106 mL-1 clearly lowered it. They also performed laparoscopic intrauterine
insemination with thawed semen, but they did not achieve significant figures. That
study showed that freezing ram spermatozoa at concentrations much higher than those
used as standard (800×106 mL-1) could be detrimental. However, these authors did not
reach to definitive conclusions, possibly due to the lack of power in their analyses and
to the presence of confounding factors. Although their results suggest a negative effect
of increasing sperm concentrations, that trend was not clear. Several studies in different
species support this hypothesis. Nascimento et al.  evaluated stallion semen doses
frozen at different concentrations: 100, 200 and 400×106 mL-1, in 0.5 ml and 0.25 ml
straws. Those authors found that sperm motility decreased with sperm concentration.
Similarly, Peñan and Linde-Forsberg , evaluated the effect of freezing dog semen at
four different sperm concentrations (50, 100, 200, and 400×106 mL-1), in 0.5 ml straws
finding sperm motility and viability after thawing was significantly lower in samples
frozen at 400×106 mL-1.
Increasing the sperm concentration might improve vaginal AI in sheep, by allowing
more spermatozoa per dose. Paradoxically, this increase could drive to the opposite
effect, if high sperm concentration at freezing would decrease sperm quality. Therefore,
we aim at confirming and improving D'Alessandro findings. It is important to confirm
and enhance these findings, in order to improve sheep AI. Thus, the objective of this
study is to assess the post-thawing sperm quality and fertility of ram semen frozen in
different concentrations (200, 400, 800 and 1600×106 mL-1) with a possible practical
use for AI in sheep. In this study we have tried to avoid confounding factors ?
equalizing the number of spermatozoa inseminated?, and we have used sensitive
techniques (CASA and flow cytometry), in order to reach more definitive conclusions,
and the fertility study was carried out using sheep groups large enough to attain a high
2. Materials and methods.
Reagents were obtained from Sigma (Madrid, Spain), except fluorescence probes
SYBR-14 (LIVE/DEAD Sperm Viability Kit) and YO-PRO-1, which were acquired
from Invitrogen (Barcelona, Spain).
2.2. Animals and sperm collection.
We used seven adult males (2-9 years old) of the Churra breed, of proven fertility and
trained for semen collection by artificial vagina. Ejaculates were collected by artificial
vagina at 40 °C (Minitüb, Tiefenbach, Germany), and the tubes were maintained at
35 °C during the initial evaluation of semen quality. The volume was estimated by using
the graduation marks of the collection tube. Mass motility was assessed by microscopy
(warming stage at 37 °C, ×40; score: 0-5; Labophot 2, Nikon, Tokyo, Japan), and the
sperm concentration was assessed by the photocolorimetric method at 540 nm
(Spectronic 20, Baush & Lomb, Madrid, Spain), on a specific calibrated scale. Only
The seven males yielded 18 good-quality ejaculates, which were divided into four
aliquots and frozen at 4 different sperm concentrations (200, 400, 800 and
1600×106 mL-1), obtaining a total of 679 straws. Semen collection was performed from
September to November (within the breeding season, which spans from July to
December). Four males yielded three good-quality ejaculates, whereas and the
remaining three yielded two good-quality ejaculates.
2.3. Cryopreservation protocol
Semen was diluted with the same volume (1:1) of freezing extender. The freezing
extender was of own design (UL) , consisting of a TTF medium (TES-Tris-fructose,
320 mOsm/kg, pH 7.2) supplemented with 10% egg yolk and 4% glycerol. The sample
was then refrigerated in a cold room at 5 °C for an average of two hours, until the
samples reached a temperature of 5 °C. At that point, the sample was divided among
four tubes, to whom more extender was added to obtain a concentration of 1600, 800,
400 or 200×106 sperm/ml. Samples were packed into 0.25-mL plastic straws and
equilibrated for 1 h at 5 °C. Then, the straws were frozen using a programmable
biofreezer (Kryo 10 Series III; Planer plc., Sunbury-On-Thames, UK) using a rate of -
20 K/min down to -100 °C. The straws were kept in liquid nitrogen containers and
stored for a minimum of two months until analysis. Thawing was carried out in a water
bath at 65 °C for six seconds. Sperm quality parameters were evaluated immediately
2.4. Spermatozoa evaluation.
The assessment of motility parameters was carried out using a computer-assisted sperm
analysis system (CASA) (ISAS v. 1.1; Proiser, Valencia, Spain). Samples were diluted
(10?20×106 cells/ml) in the same TTF medium with 320 mOsm/kg, and warmed on a
37 °C plate for 5 min. Then, a 5-µL drop was placed into a Makler counting cell
chamber (10 µm depth; Sefi Medical Instruments, Haifa, Israel). The sample was
examined at ×10 (negative phase contrast) in a microscope (Eclipse E400, Nikon) with
a warmed stage (38 °C). The standard parameter settings were set at 25 frames/s, 20 to
90 ?m2 for head area and VCL > 10 ?m/s to classify a spermatozoon as motile . At
least five sequences or 200 spermatozoa were saved and analyzed afterwards. Reported
parameters were curvilinear velocity (VCL, µm/s), linearity (LIN, %), and amplitude of
lateral head displacement (ALH, µm). Total motility (TM) was defined as the
percentage of spermatozoa with VCL > 10 µm/s, and progressive motility (PM) was
defined as the percentage of spermatozoa with VCL > 25 µm/s and STR > 80%
(straightness, also provided by the system).
2.5. Sperm viability and acrosome status.
Viability and acrosomal status were assessed simultaneously using florescence probes
and flow cytometry, according to methods described previously . Briefly, samples
were diluted in PBS at 5×106 spermatozoa/ml, and incubated for 15 min with 24 µM of
propidium iodide (PI) and 1 µg/ml of PNA-FITC (peanut agglutinin). PI stains
membrane-damaged spermatozoa red, whereas PNA-FITC stains the acrosome green if
it is damaged or reacted. Thus, we obtained four different subpopulations: red (non-
viable sperm, intact acrosome), green (viable sperm, damaged acrosome) red and green
(non-viable sperm, damaged acrosome) or not-stained (viable sperm, intact acrosome).
As a caveat, the PNA-FITC stain may have a low percentage of false-negatives, since
spermatozoa with a completely lost acrosome cannot be stained by PNA-FITC. The
percentage of spermatozoa with damaged acrosomes (ACR) was calculated as the sum
of viable and non-viable PNA+ spermatozoa.
To evaluate sperm viability, we used the double stain SYBR-14/PI. Sperm samples were
diluted with PBS down 5x106 sperm/ml, and incubated for with 24 µM PI and 100 nM
SYBR-14. The tubes were kept at 37 °C for 20 min in the dark. We detected three
populations corresponding to live spermatozoa (green), moribund spermatozoa (red +
green) and dead spermatozoa (red).
YO-PRO-1/PI was used distinguish three populations of sperm: sperm nucleus with red
fluorescence (PI+, dead), spermatozoa with green nucleus indicating intracellular
YO-PRO-1 (increased membrane permeability) and unstained spermatozoa (viable).
The diluted sample was stained with 100 nM of YO-PRO-1 and 24 µM of PI, and then
incubated at 37 °C for ten minutes before being analyzed by flow cytometry. In this
analysis, we also calculated the ratio (RATIO) among the proportion of spermatozoa
with increased membrane permeability (PI-/YO-PRO-1+) and the proportion of PI-
spermatozoa (sum of YO-PRO-1- and YO-PRO-1+).
Evaluation of flow cytometer parameters was carried out using a FACScalibur flow
citometer (Becton Dicknson System, San Jose, CA, USA) equipped with standard optics
and an argon-ion laser, tuned at 488 nm, and running at 200 mV. Calibration was carried
out periodically using standard bead (Calibrites: Becton Dickson). Data corresponding
to the red (FL-3 photodetector) and green (FL-1 photodetector) fluorescence of 10,000
spermatozoa were recorded for each stain combination.
2.6. Insemination procedures.
A total 762 adult Churra ewes were used and distributed into three experimental groups
(200, 400 and 800×106 mL-1) during the breeding season. The 1600×106 mL-1 treatment
showed a clear detrimental effect in the in vitro tests, and it was not included in the
fertility trials. The ewe number was estimated through a power analysis, taking into
account D'Alessandro et al.  results, to detect a difference of at least 13 points in
fertility rates, with a statistical power of 0.9 and a signification level of 0.05 (total
number of females: 745). These females were subjected to treatment for oestrus
induction and synchronization using intravaginal sponges with 40 mg fluorogestone
acetate for 14 days. Then, the sponges were removed and the ewes were treated with
500 IU of eCG (i.m.). Laparoscopic inseminations were performed by two experienced
technicians between 62 and 64 hours after the removal of the sponges. The animals,
having fasted for the previous 24 hours, were placed on a special cradle (IMV®)
adjusted at an inclined plane (45°). The abdominal area in front was shaved and cleaned.
Then, two portals (for vision and manipulation/injection) were inserted by performing a
pneumoperitoneum (CO2). The semen, placed in a special applicator (transcap, IMV®),
was injected under visual inspection into both uterine horns (0.12 mL per horn). Sperm
concentration was equalized to 200×106 mL-1 just before insemination, using freezing
extender, thus 25×106 spermatozoa were applied per horn in the three treatments.
Fertility results were noted as lambing rates (percentage of lambing ewes at 137-154
day post-insemination respect to the total number of inseminated ewes).
2.6. Statistical analysis.
Statistical analyses were carried out using the R statistical package, version 2.13.0
(http://www.r-project.org). Data were fitted to linear mixed-effect models (lmer
package) by maximizing the log-likelihood (ML method) . Sperm concentration
(four levels) was included in the fixed part of the model, whereas male and ejaculate
within male were included in the random part of the model. A pairwise comparison
among sperm concentrations was performed whenever the effect of sperm concentration
was significant, using Tukey contrasts. Fertility results were analyzed by logistic
regression. Odd ratios (OR) and 95% confidence intervals (CI) were generated during
the logistic regression. Results are given as mean±SEM.
The results of this experiment are showed in the Table 1. The CASA parameters TM,
PM and VCL did not show significant differences between the concentrations 200, 400
and 800×106 mL-1, but the highest concentration used in this experiment
(1600×106 mL-1) yielded significantly lower results (TM and PM: P<0.001 for 200, 400
and 800 vs. 1600; VCL: P=0.024 for 200 vs. 1600, P=0.002 for 400 vs 1600, and
P<0.001 for 800 vs. 1600). Mean values of LIN and ALH were not significantly
different among concentrations.
The analysis of physiological parameters using fluorescence probes showed that the
highest concentration (1600×106 mL-1) yielded significantly lower viability (figures 1, 2
and 3) and a higher "apoptotic ratio" (Figure 3), comparing to the other concentrations.
Contrarily, the proportion of damaged acrosomes (ACR) was little affected by sperm
concentration, with no significant differences detected among concentrations (Figure 1).
The proportions of the subpopulations obtained from the SYBR-14/PI and YO-PRO-
1/PI stains varied among concentrations, as showed in figures 2 and 3. The viable
subpopulation according to both stains showed its highest value for 400×106 mL-1,
being 800 and 1600×106 mL-1 significantly lower, especially the latter. Interestingly,
while the percentages of dead (PI+) spermatozoa showed little changes among
concentrations (being significantly higher for 1600×106 mL-1), the percentages of
moribund (SYBR-14+/PI+) and spermatozoa with increased membrane permeability
(YO-PRO-1+/PI-) increased significantly when freezing at 800×106 mL-1
(200×106 mL-1, P=0.048; 400×106 mL-1, P=0.036) and 1600×106 mL-1 (P<0.001
comparing with the other three concentrations).
The fertility results after intrauterine insemination (lambing rates) are reported in the
Table 2. Fertility was affected by the sperm concentration used for freezing the semen
doses. It was significantly higher when the ewes were inseminated with doses frozen at
200 and 400×106 mL-1 (yielding 57.5 and 54.4% fertility, respectively), comparing with
800×106 mL-1 (45.5%; P<0.05). Females inseminated with 800×106 mL-1 doses were
0.62 times less likely of getting pregnant than those inseminated with 200×106 mL-1
doses (odds ratio, 95% CI: 0.44?0.88; P=0.007).
Semen cryopreservation induces a series of structural and biochemical changes in
spermatozoa, thus reducing the integrity of the membrane [18,19], mobility [13,14] and
fertilizing capacity [20,21]. Many factors influence the survival and functionality of the
frozen/thawed spermatozoa, but sperm concentration has been little explored. The
optimization of semen doses and the utilization of sorting technologies has driven
interest towards freezing using low sperm concentrations [22,23]. However, freezing at
high concentrations might be interesting in some cases, and it could be necessary to
increase the absolute number of fertile spermatozoa post-thawing while still managing a
small volume. Utilization of very high concentrations is usual in fish species, and it
does not seem to be detrimental for the cryopreservation of spermatozoa . However,
increasing sperm concentration for freezing could have undesirable consequences in
mammals, exceeding the possible advantages of this approach.
Our study follows D'Alessandro et al. . Recapitulating from the introduction, these
authors evaluated the survival of sheep semen frozen at several concentrations (50, 100,
200, 400, 500 and 800×106 mL-1) and in two extenders (egg yolk-based ?Tris-FY?
and milk-based ?Milk-LY). Irrespective of the extender, they found that the overall
performance (motility, viability and acrosomal status) decreased when freezing at
800×106 mL-1. Nevertheless, results were not entirely conclusive, with a high variability
among concentrations. Subjective motility was lower in the 800×106 mL-1 samples. In
our results, the progressive motility was significantly different among 200 and
800×106 mL-1, although total motility and the kinematic parameters remained similar.
Viability also decreased in that study for 800×106 mL-1, although they reported low
viability for several of the lower concentrations too. Interestingly, in our results
400×106 mL-1 yielded a viability higher than 800×106 mL-1, as assessed using PI/SYBR-
14 or PI/YO-PRO-1, while 200×106 mL-1 stayed in between, not being significantly
different than 800×106 mL-1.
Other studies have detected an effect of high concentrations on cryopreservation yields
in other species. Nascimento et al.  compared the motility, viability and
mitochondrial activity of stallion spermatozoa frozen at 100, 200 and 400×106 mL-1.
They found highest motilities at 200×106 mL-1, followed by 400×106 mL-1, and the
lowest values at 800×106 mL-1. Similarly, Crockett et al.  found higher post-thaw
progressive motility in cooled samples and after cryopreservation at concentrations of
50 and 250×106 mL-1 (25% and 23%, respectively) than in samples at a concentration of
500×106 mL-1 (17%). Peña and Linde-Forsberg  obtained diverging results when
testing 50, 100, 200 and 400×106 mL-1. Whereas viability was higher at lower
concentrations, there were no differences on progressive motility and, after incubating
the samples for several hours, 400×106 mL-1 yielded both the highest progressive
motility and viability. Therefore, the spermatozoa of some species, such as the horse,
seems to be sensitive to high concentrations while freezing, whereas others, dog and
possibly ovine, seems to be resilient and even being better cryopreserved at moderately
The causes of this decrease are still little know, but could be multifactorial and
intertwined: excess of free radicals, modification of the sperm metabolism, changes in
the media due to catabolism products, physical changes during the freezing/thawing,
etc. Moreover, acrosomal enzymes and toxic products released from damaged
spermatozoa (e.g., free radicals) might contribute to the destabilization of membranes
and other structures in live spermatozoa, and it could have a larger effect at higher
sperm concentration [26,27]. We aimed at exaggerating any detrimental effects by using
the 1600×106 mL-1 concentration. Indeed, the proportion of spermatozoa with damaged
membranes (loss of viability) decreased clearly in that treatment, explaining at least in
part the concomitant loss of motility. In equine spermatozoa, Crockett et al.  found
not only a lower motility, but also a higher percentage of sperm with damaged
membranes in the 500×106 mL-1 doses (45%) than in the 50×106 mL-1 doses (60%).
Interestingly, we could not detect an increase of acrosomal damage, even in the
1600×106 mL-1 samples. Nevertheless, we must take into account that the absolute
quantity of enzymes and other molecules released from damaged acrosomes (and, in
general, from dead spermatozoa) increase with sperm concentration. Therefore, at the
same proportion of damaged acrosomes, samples with higher sperm concentration
would have a higher concentration of these potentially harmful substances, which might
explain, at least in part, the lower sperm quality.
The YO-PRO-1 stain has the ability to label spermatozoa that have an increasing
membrane permeability that not necessarily implies a loss of continuity [28,29],
resembling some phenomena occurring in apoptotic somatic cells. We detected that the
proportion of spermatozoa showing these early membrane changes increased with
sperm concentration. This increase was more evident when the presence of the
???????????? ??????????? ???? ?????????? ??? ??? ??????????? ???????? ????? ??? ??? ??????????
variable, since these subtle membrane changes could announce a higher vulnerability of
inseminated spermatozoa to the oviductal environment and a lower ability to attach to
the oviductal epithelium . Therefore, samples frozen at 800 or 1600×106 mL-1 not
only had a lower proportion of viable spermatozoa, but a higher proportion of
potentially non-functional spermatozoa .
Indeed, the fertility obtained following intrauterine artificial insemination was
significantly lower in ewes inseminated with 800×106 mL-1 doses. ????????????????????
 observed that increasing the pre-freezing sperm concentration to 800×106 mL-1
negatively affected the proportion of pregnant ewes, but their results were not
significant. Moreover, inseminations were performed without equalizing sperm numbers
among different sperm concentrations, and the excess of spermatozoa when using the
doses with higher concentration could compensate in part for the lack of quality. In our
study, we used a larger number of females per insemination group, and utilized the same
number of spermatozoa per insemination, irrespectively of the dose concentration. We
have achieved an overall fertility above 50%, similarly to previous studies with
laparoscopic insemination [3,4]. The best results for fertility were achieved by 200 and
400×106 mL-1, related to the highest results for motility, plasma membrane integrity and
permeability. The 800×106 mL-1 doses obtained an odds ratio of 0.62 respect to
200×106 mL-1 doses. That is, the odds for an insemination with a 800×106 mL-1 dose to
results in a pregnancy are 0.62 times lower than for an insemination with a 200×106 mL-
It is true that pre-insemination extension might have penalized the 800×106 mL-1 doses.
However, we think that our trial is realistic, since dilution would occur in vivo after
routine insemination, penalizing these spermatozoa anyway. Moreover, extension
sensitivity might be due to sub-lethal membrane damage. Given that samples frozen at
400×106 mL-1 yielded fertility results similar to the 200×106 mL-1 doses, we can
conclude that these samples were not affected by extension, remarking that detrimental
effects occur when samples are frozen at higher concentrations.
In conclusion, increasing sperm concentration in the sperm doses (at least, above
400×106 mL-1) affects adversely the post-thawing quality and fertility of ram semen.
Sperm quality was slightly affected at 800×106 mL-1, but membrane changes
(proportion of spermatozoa with apoptotic features) indicated further detrimental effects
upon use of the doses in the field. Actually, freezing at those sperm densities affected
negatively the fertility of the samples.
One of the purposes of freezing at high concentrations is to increase the number of
spermatozoa available in the insemination dose, attempting to improve the odds for
achieving a pregnancy. Moreover, this increase of inseminated spermatozoa would
compensate for some quality decrease when freezing at high concentrations.
Nevertheless, taking into account the results of D'Alessandro et al. , it seems that
this hypothesis may not be correct, since their insemination results tended to decrease
with the 800×106 mL-1 doses, even though they inseminated with a higher number of
spermatozoa (not at a fixed number, like in our case). Therefore, trying to improve AI
results in sheep by freezing and inseminating at a higher sperm concentration might not
compensate the excess spermatozoa used. Testing this hypothesis should be an objective
in future studies. Nevertheless, it is still open to further research if there is any
advantage freezing at 400×106 mL-1 (or higher) and inseminating with the full dose.
Since freezing at 200 and 400×106 mL-1 have yielded similar performance for in vitro
quality and in the fertility trial, it is reasonable to think that ram semen could be frozen
at least at that concentration, thus increasing the odds of pregnancy.
This work was supported in part by INIA (RZ2010-00005-00-00). Felipe Martínez-
Pastor was supported by the Ramón y Cajal program (RYC-2008-02560, MICINN,
Spain). The authors thank Manuel Alvarez, María Nicolas, Susana Gomes and Elena
López for their help in the acquisition and analysis of the samples.