| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Veterinary Teaching Hospital Service of Reproduction and Obstetrics Faculty of Veterinary Medicine, University of Extremadura Avd de la Universidad s/n Cáceres, Spain.
| Correspondence to: Dr F. J. Peña, Section of Reproduction and Obstetrics Department of Herd Health and Medicine Faculty of Veterinary Medicine Avd de la Universidad s/n 10071, Cáceres, Spain (e-mail: fjuanpvega{at}unex.es). |
| Received for publication October 3, 2005; accepted for publication March 12, 2006. |
 |
Abstract |
|---|
 Top Abstract Material and MethodsResultsDiscussionReferences |
|---|
Key words: Dog, computer-assisted sperm analysis
The second step (cluster step) takes subclusters (non-outlier subclusters if outlier handling is used) resulting from the precluster step as input and then groups them into the final number of clusters. SPSS uses the agglomerative hierarchical clustering method (www.rrz.uni-hamburg.de/RRZ/Software/SPSS/Algorith.120/twostepcluster.pdf).
This system allows the user to fix the previous maximum number of clusters or let the technique automatically choose the number of clusters with either the Bayesian information criterion or Akaike information criterion. The great advantage of this system is that all the analyses can be done in 1 step more rapidly and easily than with the techniques used so far to disclose sperm subpopulations.
Another problem derived from the analysis of the data originating from the CASA-ASMA analysis is the large variety of separated parameters that in most cases are highly correlated. When there are several correlated variables in biological systems, principal component analysis (PCA) can be used to reduce these to 1 or 2 variables that are linear functions of the original variables (Agarwal et al, 2003). Therefore, the aims of this study were to:
|
Material and Methods |
|---|
|
TopAbstractMaterial and MethodsResultsDiscussionReferences |
|---|
Experimental Design
Five ejaculates were obtained from each dog once a week for 5 consecutive
weeks. After collection, an aliquot was removed for computer-assisted
evaluation of motility and morphometry. After 4 weeks of storage, semen was
thawed and sperm motility was evaluated again.
Semen Collection
Semen was collected by masturbation in a prewarmed graduated test tube.
After collection, sperm samples were kept at 37°C in a water bath. An
aliquot was removed for sperm concentration measurement and evaluation of
motility and morphology (phase contrast, microscopy). Only samples with at
least 70% motility and 80% normal morphology were included in the study.
Semen Processing
Semen was processed by the Uppsala method
(Peña and Linde Forsberg,
2000) with modifications. In brief, after collection semen was
diluted 1:1 in Tris-glucose extender I (200 mM Tris, glucose 70 mM, citric
acid 63 mM, bovine serum albumin 3%, penicillin 1000 UI/mL,
dihydroestreptomicin 1 mg/mL) and centrifuged for 8 minutes at 700 x
g. The seminal plasma was then removed and the sperm pellet was
resuspended in extender II at room temperature (200 mM Tris, 70 mM glucose, 63
mM citric acid, glycerol 3%, vol/vol egg yolk 20%, vol/vol penicillin 1000
UI/mL, dihydroestreptomicin 1 mg/mL) and cooled to 5°C over a period of 1
hour. After equilibration, an equal volume of Extender III (200 mM Tris, 70 mM
glucose, 63 mM citric acid, 7% vol/vol glycerol, 20% vol/vol egg yolk, 1%
vol/vol Equex STM paste [Nova Chemical Sales Inc, Sciutate, Mass] penicillin
dihydroestreptomicin) was added at 5°C to a final sperm concentration of
150 to 200 x 106 spermatozoa per milliliter. The sperm was
then loaded in 0.5-mL straws and frozen horizontally in racks, placed 4 cm
above of the surface of liquid nitrogen in a closed Styrofoam box for 10
minutes, and then plunged directly in liquid nitrogen. After 4 weeks of
storage, the straws were thawed in a water bath at 70°C for 8 seconds.
Motility Analysis
Motility was measured in fresh samples and after cryopreservation by using
a CASA system (ISAS, Proiser SL, Valencia, Spain). Analysis was based on the
examination of 25 consecutive, digitalized images obtained from a single field
with a 10x negative-phase contrast objective. Images were taken with a
time lapse of 1 second; the image capture speed was therefore 1 every 40
milliseconds. The number of objects incorrectly identified as spermatozoa was
minimized on the monitor by using the playback function. With respect to the
setting parameters for the program, objects with a curvilinear velocity less
than 10 µm/s were considered immobile, whereas objects with a velocity
greater than 15 µm/s were considered motile. Objects with velocities
between 65 and 100 µm/s were understood as medium-speed objects, whereas
those with a velocity greater than 100 µm/s were considered rapid objects.
Spermatozoa deviating less than 10% from a straight line were designated
linear motile. Sperm motion kinematics measured by CASA included the
following:
|
Sperm Staining for Morphometric Analysis
Before ASMA analysis, sperm samples were washed once in phosphate-buffered
saline (PBS) by centrifugation at 800 x g for 5 minutes and were
adjusted in PBS to 100 x 106 cells per milliliter. Then, 10
µL of the sperm suspension was placed on the edge of a slide and extended.
Preparations were allowed to dry and were then fixed and stained for 10
minutes in an eosin solution (Panreac, Barcelona, Spain) and for 10 minutes in
a methylene blue solution (Panreac). The excess of staining was removed, and
the slide was allowed to dry and was permanently mounted (Eukitt, Panreac)
Computerized Morphometric Analysis
The prepared slides were examined with a Nikon Labophot microscope equipped
with a 100x bright field objective and a 3.3x photo-ocular. The
video signal was acquired with a Sony CCD AVC-D7CE video camera (Sony
Corporation, Tokyo, Japan) interfaced with an ISAS CASMA system (Proiser SL).
The array size of the video grabber was 512 x 512 x 8 bits
providing digitized images of 262.144 pixels and 256 gray levels. Resolution
of images was 0.083 µm per pixel in the horizontal and vertical axes.
Sperm cells were displayed on the monitor at equivalent brightness, and all
the cells that did not present any overlap with debris or other cells were
considered for analysis. From each sample, heads were captured and analyzed by
the program as previously described
(Buendia et al, 2002). After
treatment of the images, some of the cells had to be discarded because of
defective binarization as observed by false correspondence between the
original image and its mask. Each sperm head was measured for 9 primary
parameters (head area [A], µm2; head perimeter [P], µm; head
length [L], µm; head width [W] µm; percentage of the sperm head occupied
by the acrosome; midpiece width [w], µm; midpiece area [a], µm; distance
[d] between the major axes of the head and midpiece, µm; angle [h] of
divergence of the midpiece from the head axis) and 4 derived parameters of
head shape (FUN1 [L/W], FUN2 [4
A/P2], FUN3 [(L - W)/(L + W)],
FUN 4 [LW/4A]).
Statistical Analysis
The data matrix consisted of 3446 observations for fresh semen and 5773 for
frozen-thawed samples for the motility analysis, whereas 2361 individual
spermatozoa were evaluated for the morphometric analysis.
The main objective of the analysis was to extract sperm subpopulations by
using the data obtained from each dog by means of clustering procedures. The
level of significance was set at P< .05. The first step was to
perform a PCA of the data. The purpose of the first step was to derive a small
number of linear combinations (principal components [PCs]) that retain as much
of the information in the original variables as possible. This allows one to
summarize many variables in few, jointly uncorrelated PCs. A good result is
considered if we obtain a few PCs accounting for a high proportion of the
total variance. The VARIMAX method with Kaiser normalization was used as a
rotation method. The second step was to perform a 2-step cluster procedure
with the sperm-derived indexes obtained after the PCA. To study the
distributions of observations (individual spermatozoa) within dogs and
ejaculates and within subpopulations, we used the analyses of variance and
2 tests. Linear regression analyses were used to investigate
relationships among sperm indexes and sperm velocities postthaw. For
comparison of mean values in fresh and frozen-thawed semen, the GLM procedure
followed by Tukey's test was used. All analyses were performed with SPSS
version 12.0 for Windows software.
Calculation of Sperm CASA-Derived Indexes
We used a simplified version of the technique described by Agarwal et al
(2003). We first performed a
PCA and then weighed variables against their eigen vectors. Using this
approach, we proposed the following formulas:
 |
 |
 |
 |
 |
 |
|
Results |
|---|
|
TopAbstractMaterial and MethodsResultsDiscussionReferences |
|---|
|
Development of Sperm Quality Indexes as Measures of Semen Quality
PCA in CASA data revealed that 2 components accounted for more of the 91%
of the cumulative variance. The first component was related to sperm
velocities (SVI) and also included ALH. The second component was more related
to the characteristics of the sperm movement itself (SMI). Because these
scores are produced by PCA, SVI and SMI are not correlated and resume all the
information obtained after the CASA. In this way, VCL, VSL, VAP, ALH, STR,
WOB, and LIN are resumed in SVI and SMI.
For the morphometric values, 4 PCs accounted for more of the 70% of the
variance. The first 2 PCs were related to sperm head size and shape, whereas
PCs 3 and 4 were mainly related to characteristics of the midpiece. These 4
sperm indexes resumed 12 ASMA-derived sperm morphometric
variables.
Sperm Subpopulations Based on Kinematics Properties of the Sperm
The 2-step clustering procedure revealed the existence of 6 subpopulations.
Subpopulations 4 and 6 were characterized by high values of both SVI and SMI,
subpopulations 2 and 3 were characterized by medium values, and subpopulations
1 and 5 were characterized by low values
(Table 1). The distribution of
sperm subpopulations was completely different among dogs
(Table 2). Two sperm
subpopulations were present in each dog: subpopulations 5 and 6 were in dog 1,
subpopulations 3 and 4 were in dog 3, subpopulations 1 and 4 were in dog 4,
and subpopulations 1 and 4 were in dog 5.
|
|
Sperm Subpopulations Based on Morphometric Properties of the Spermatozoa
Four sperm subpopulations based on morphometric parameters of the sperm
head and midpiece were revealed (Table
3). Also, the distribution of sperm subpopulations was completely
different in each dog (Table
4).
|
|
Relationship Among Sperm-Derived Indexes and Sperm Quality Postthaw
VSL, VAP, and VCL were used as indicators of sperm quality postthaw.
Stepwise linear regression analyses using the sperm indexes as predictive
variables were used to forecast these 3 sperm velocities separately. We choose
these variables because they are widely considered to be good indicators of
sperm quality. In addition, there are a number of studies showing a high value
of VCL in the prediction of human sperm quality
(de Geyter et al, 1998;
Larsen et al, 2000) and field
fertility in animal models (Holt et al
1997). Thus, the maintenance of high velocities postthaw may be a
good indicator of the success of cryopreservation.
The 2 models explained VCL postthaw; however, the residuals sum of squares was much lower in model 2. This model had R2 = 0.997, which explained the 99% in the variation. The model included a constant, SVI, and SMI (Table 5) and was statistically significant (P < .001).
|
Two models were developed for VSL and VAP; the ones including both SVI and SMI index were elected because of the lower values of the residual sum of squares. However, this was much bigger in the case of VSL than VAP.
The model predicting values of VAP had R2 = 0.98 (P < .001), and the model predicting values of VSL had R2 = 0.99 (P < .0001) (Tables 6 and 7).
|
|
|
Discussion |
|---|
|
TopAbstractMaterial and MethodsResultsDiscussionReferences |
|---|
From this evidence, it is becoming clear that the classical approach considering the ejaculate as a homogeneous population following a normal statistical distribution is no longer valid.
In this study, we used simple indexes resuming all the variables of the CASA for morphometric and kinematics data separately, and we used these data to disclose sperm subpopulations. Using this approach, we were able to identify 6 sperm subpopulations by kinematics data and 4 sperm subpopulations by morphometric data. Given the new information regarding the quality of individual ejaculates, the distribution of clusters varied significantly among dogs. Classical analysis of sperm data by using mean values did not reveal differences among dogs.
Also, the approach used in our study allowed us to predict the freezeability of a semen sample. We used as a measure of freezeability the changes in sperm velocities after freeze-thaw procedures, for it is assumed that these parameters are a manifestation of functional competence of spermatozoa and they are normally positively correlated with those of plasma membrane and sperm morphology (Peña, 2004). Also, a number of studies show that sperm velocities are sound indicators of sperm quality (de Geyter et al, 1998; Larsen et al, 2000). The predominance of a particular subpopulation in samples with better cryoresistance or fertility has been demonstrated in other studies (Quintero-Moreno et al, 2003; Martinez-Pastor et al, 2005; Peña et al, 2005).
The results from our study strongly support the hypothesis that the mammalian ejaculate is a heterogeneous cellular suspension and that the erroneous use of data generated by the CASA was attributed to frequently disappointing results when motility was used as an indicator of sperm quality (Abaigar et al, 1999; Quintero-Moreno et al, 2003).
It is clear that not all the spermatozoa in an ejaculate have the same biochemical status and thus respond in a different manner to different stimuli (Abaigar et al, 1999). It is also clear that not all the spermatozoa have the same morphometry (Peña et al, 2005), and thus the cell surface area is different (Mazur and Koshimoto, 2002), thereby affecting the exchanges of heat, water, and ions during cryopreservation in a different manner in each subpopulation (Peña et al, 2005). Clearly, the subpopulations structure approach has advantages over the use of mean values for the study of the ejaculate; the former identifies subpopulations of spermatozoa in different biochemical status, whereas the use of mean values just masks it (Abaigar et al, 1999) and also can identify morphometric subpopulations with a cell surface area more adequate to support the stress of cryopreservation (Peña et al, 2005). Also, the use of combined indexes by PCA can better reflect the energetic status of each subpopulation. In our study, SMI and SVI were able to predict sperm velocities postthaw. The statistical technique is advantageous because of its simplicity; whereas previously described techniques imply 2 or 3 steps (Abaigar et al, 1999; Martinez-Pastor et al, 2005), all the clustering procedures here are performed in 1 step. Also, the system manages outliers automatically. This statistical tool proved to be very powerful when disclosing sperm subpopulations despite the relative low number of dogs used in this study. Recent studies (Rathi et al, 2001; Quintero-Moreno et al, 2003) have suggested that studies with a relatively low number of animals can produce very interesting proposals for the study of the mammalian ejaculate if the data are adequately processed.
Interesting findings of our study are those related to the predictability of sperm cryoresistance. As stated earlier, median values of sperm kinematic parameters did not differ among dogs; however, the pattern in the distribution of sperm subpopulations varied. It is noteworthy to remark that this interpretation would not have been possible without considering sperm subpopulations. The dog with the worst ability to support cryopreservation (dog 3) had predominated sperm subpopulations with low values of SVI and SMI in fresh samples. On the other hand, the dogs showing the best cryoresistance (dogs 1 and 5) had predominated sperm subpopulations with higher values of SMI and SVI. Also, it is noteworthy that dog 1 showed higher cryoresistance. In this dog, there was only a slight decrease in the percentage of progressive motile sperm after freeze-thaw (Figure) and a clear predominance of kinematic cluster 6, characterized by the highest values of SVI and very high values of SMI. It is plausible that these sperm indexes resume the biological status of sperm cells. It is interesting to note that different sperm velocities (VCL, VAP, VSL) are regulated in a different manner, as demonstrated by different tyrosine kinase inhibitors (Bajpai and Doncel, 2003). Therefore, these indexes may be a reliable indicator of the biochemical status of the sperm subpopulation and thus may be used to predict cryoresistance. This may be especially relevant in species showing a relatively high cryoresistance, such as dogs. In fact, from our study this theory may explain the high predictive value of this approach. By using the PCA-derived indexes, we can resume in few variables the complex structure and biochemical status of the ejaculate. Those ejaculates with a predominance of subpopulations with high SMI values will show higher sperm velocities after freezing. If we evaluate what happened in dog 5 (the dog with higher sperm velocities after freezing) kinematic clusters 2 (mid to high values of SVI and SMI) and 4 (very high values of SMI and SVI) predominated. It is very plausible to think that most of the sperm not surviving the freeze-thaw procedure belonged to cluster 2; thus, in this dog there was not a marked decrease in sperm velocities after thawing. Following this reasoning, we can also explain why dog 3 had the worst cryoresistance; in this dog, more than the 90% of spermatozoa belonged to a cluster with low values of SVI and SMI. It is very important to note that when using mean values of sperm motility we were unable to detect statistically significant differences among dogs. Cryopreservation implies much stress to the spermatozoa (Mazur, 1984). Perhaps the 2 main types of stress are osmotic stress and the formation or reshaping of intracellular ice during freezing and again during thawing. In more sensitive species such as pigs, it may be plausible that most of the cryoinjury is physical, whereas in most cryoresistant species such as dogs, the biochemical injury may be more present in the form of subtle damages of the sperm membranes, resulting in alterations of the cell physiology. Our results support this theory, because rather than a high percentage of inmotile sperm after thawing we observed mainly decreases in the percentage of progressive motile spermatozoa and decreases in all sperm velocities.
In conclusion, we have demonstrated that a simple 2-step cluster procedure can be used to disclose sperm subpopulations within a semen sample. This system is easier than the previously used statistical procedures because it allows 1-step entry of data and the automatic management of outliers. Also, we have demonstrated that the use of PCA-derived indexes can be used as variables to enter in the cluster procedure, retaining most of the information of the original variables and much more useful information regarding the characteristics of the ejaculates and its freezeability than provided by conventional semen analysis. Finally, we propose the use of a multivariate clustering procedure as standard analysis of semen samples when CASA systems are used.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractMaterial and MethodsResultsDiscussionReferences |
|---|
Agarwal A, Sharma RK, Nelson DR. New semen quality scores developed
by principal component analysis of semen characteristics. J
Androl. 2003;24: 343
–352.
Bajpai M, Doncel GF. Involvement of tyrosine kinase and camp dependent kinase cross-talk in the regulation of human sperm motility. Reproduction. 2003; 126: 183 –195.[Abstract]
Buendia P, Soler C, Paolicchi F, Gago G, Urquieta B, Pérez Sánchez F, Bustos Obregón E. Morphometric characterization and classification of alpaca sperm heads using the sperm class analyzer system. Theriogenology. 2002; 57: 1207 –1218.[CrossRef][Medline]
de Geyter Ch, de Geiter M, Koppers B, Nieschlag. Diagnostic
accuracy of computer assisted sperm motion analysis. Hum
Reprod. 1998;13: 2512
–2520.
ESHRE Andrology Special Interest Group. Guidelines on the
application of CASA technology in the analysis of spermatozoa. Hum
Reprod. 1998;13: 142
–145.
Holt C, Holt WV, Moore HDM, Reed HCB, Curnock RM. Objectively
measured boar sperm motility parameters correlate with the outcomes of on farm
inseminations: results of two fertility trials. J
Androl. 1997;18: 312
–323.
Holt W. Can we predict fertility rates? Making sense of sperm motility. Reprod Domest Anim. 1996; 31: 17 –24.[Medline]
Holt WV, Van Look KJW. Concepts in sperm heterogeneity, sperm
selection and sperm competition as biological foundations for laboratory tests
of semen quality. Reproduction. 2004; 127: 527
–535.
Larsen L, Scheike T, Kold Jensen L, Peter Bonde J, Ernst E, Henrik
Hjollund N, Zhou Y, Erik Skakkebaek N, Giwercman A. Danish first pregnancy
planner study team. Computer assisted semen analysis parameters as predictors
for fertility of men from the general population. Hum
Reprod. 2000;15: 1562
–1567.
Martinez-Pastor F, García-Macias V, Alvarez M,
Herráez P, Anel L, de Paz P. Sperm subpopulation in Iberian red deer
epidydimal sperm and their changes through the cryopreservation process.
Biol Reprod. 2005; 72: 316
–327.
Mazur P. Freezing of living cells: mechanism and implications. Am J Physiol (Cell Physiol). 1984; 16: C125 –C142.
Mazur P, Koshimoto CH. Is intracellular ice formation the cause of
death of mouse sperm frozen at high cooling rates? Biol
Reprod. 2002;66: 1485
–1490.
Mortimer ST. CASA-practical aspects. J Androl. 2000;21: 514 –524.
Peña A, Linde-Forsberg CB. Effect of spermatozoal concentration and post-thaw dilution rate on survival after thawing of dog spermatozoa. Theriogenology. 2000; 54: 703 –718.[CrossRef][Medline]
Peña AI. Canine fresh and cryopreserved evaluation. Anim Reprod Sci. 2004; 82/83: 209 –224.
Peña FJ, Saravia F, García Herreros M,
Nuñez-Martinez I, Tapia JA, Johannisson A, Wallgren M, Rodriguez
Martinez H. Identification of sperm morphometric subpopulations in two
different portions of the boar ejaculate and its relation to post thaw
quality. J Androl. 2005; 26: 716
–723.
Quintero-Moreno A, Miró J, Rigau T, Rodríguez Gil JE. Identification of sperm subpopulations with specific motility characteristics in stallion ejaculates. Theriogenology. 2003; 59: 1973 –1990.[CrossRef][Medline]
Rathi R, Colenbrander B, Bevers BB, Gadella BM. Evaluation of in
vitro capacitation of stallion spermatozoa. Biol
Reprod. 2001;65: 462
–470.
Rodriguez Martinez H. Laboratory assessment and prediction of fertility: still utopia? Reprod Domest Anim. 2003; 38: 312 –318.[CrossRef][Medline]
Verstegen J, Iguer-Ouada M, Onclin K. Computer assisted semen analyzers in andrology research and veterinary practice. Theriogenology. 2002; 57: 149 –179.[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
