Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
SEMEN SEXING
The present invention relates to a method o
sorting living spermatozoa, and, for exa~ple, to a method
of sorting living spermatozoa according to sex; that
is, according to whether the spermatozoa bear an X
or Y chromosome.
Throughout the following description, the
lower case letters in parentheses refer to the
following:
(a) Almquist, J.O., Flipse, R.J. & ~hacker,
D.L. (1954) Diluters of bovine semen,
IV. Fertility of bovine spermatozoa
in heated homogenized milk and skimmed
milk. J. Dairy Sci. 37 1303-1304.
(b) Dean, P.N.,Pinkel, D. & Mendelson,
M.L. (1978) Hydrodynamic orientation
of sperm heads for flow cytometry.
Biophys. J.23, 7-13.
(c) First, N.L. (1971) Collection and
preservation of sperm. In Methods
in Mammalian Embryology p. 15-35
Ed. J.C. Daniel, Jr. Freeman, San
! Francisco.
(d) Fulwyler, M.J. (1977) Hydrodynamic
orientation oE cells. J. Histochem.
Cytochem. 25, 781-783.
(e) Gledhill, B.L., Lake, S. & Dean, P.~.
(1979) Flow cytometry and sorting
of sperm and other male germ cells.
~0 In Flow Cytometry and Sorting, pp.
--2--
471-485. Eds M.R. Melamed. P.F. Mullaney
~ M.L. Mendelsohn. Wiley, New York.
(f) Herzenberg, L.A. & Herzenberg, L.A.
(1978) Analysis and separation using
the fluorescence activated cell sorter.
In Handbook of Experimental Immunology
3rd edn, pp. 22.1 - 22.21. Ed. D.M.
Weir. Blackwell Scientific Publications,
Oxford.
(g) Herzenberg, L.A., Sweet, R.G., & Herzenberg,
L.A. (1976) Fluorescence activated
cell sorting. Sci. Am. 234, 108-117.
~h) Klasen, M. & Schmidt, M. (1981) An
improved method for Y body identification
and confirmation of a high incidence
of YY sperm nuclei. Hum. Genet. 58,
155-161.
(i) Loken, M.R., Parks, D.R. & Herzenberg,
I,.A. (1977) Identification of cell
asymmetry and orientation by light
scattering. J. Histochem. Cytochem.
25, 790-795.
(j) Lydon, M.J., Keeler, K.D. & Thomas,
D.B. (1980) Vital DNA staining and
cell sorting by flow micro-fluorometry.
J. cell. Physiol. 102. 175-181.
(k) Muller, W. & Gautier, F. (1975) Interaction
of heteroaromatic compounds with nucleic
acid A-I' specific non-intercalating
DNA ligands. Eur. J. Biochem. 54,
385-394.
(l) Russell, W.C., Newman, C. & Williamson,
D.H. (1975) A simple cytochemical
technique for demonstration of DNA
in cells infected with my~plasma
and viruses. Nature, Lond. 253,
46~-~6~.
`\
--3--
(m) Stovel, R.T., Sweet, R.G. & Herzenberg,
L.A. (1978) A means for orienting
flat cells in Elow systems. Biophys.
J. 23, 1-5,
(n) Szabo, G., Jr, Kiss, A. & Damjanovich,
S. (1981) Flow cytometric analysis
of the uptake of Hoechst 33342 dye
by human lymphocytes. Cytometry 2.
20-23~
(o) Tobey, R.A. & Crissman, H.A. (1975)
Unique techniques for microfluorometry.
Expl. Cell Res. ~3, 235-239,
(p) Van Dilla, M.A., Gledhill, B.L, Lake,
S., Dean, P.N., Gray, J.W., Kachel,
V., Barlogie, B. & Gohde, W. (1977)
Measurement of mammalian sperm deoxy-
ribonucleic acid by flow cytometry.
Problems and approaches. J. Histochem.
Cytochem. 25, 763-773.
Flow microfluorometry is a convenient method
for measuring the DNA content of mammalian cells
(o). Spermatozoa, by virtue of their ease of
collection from many species, their homogeneity
and their haploidy, are particularly suitable
for such studies (p;e). To date, the majority
of studies of the DNA content of spermatozoa have
been carried out using fixed material stained
with fluorochromes such as acridine orange, ethidium
bromide, or mithramycin Recently, t~ bisben-
zimidazole dyes Hoechst 33258, Hoechs 3342,and DAPI(4',6'-diamidino-2-phenylindole) have
been introduced as quantitative fluorescent stains
for DNA. These dyes, although they bind tightly
to DNA, do not intercalate into the molecule and
hence are reputed not to disrupt its structure
(k;l). ~hese fluorochrome dyes are consequently
capable of being used as quantitativr vital stains
~'25~
for DNA: Hoechst 33258 and Hoechst 33342 have been used
as vital stains to distinguish phases of the cell cycle.
Since spermatozoa are tail bearing and motile they
orisntate with their long axis zlong the line of flow in
a flow microfluorometry system (p). It has been
concluded that an apparent bimodal DNA distribution in
fixed acriflavine/Feulgen-stained bull sperm heads
analyzed in such a system, is due to an orientation
artefact (b)/ perhaps analogous to that previously
described in (i) for the light scatter (size) artefact
seen with chicken red blood cells (chicken RBC). Both
of these artefacts can be by-passed ox removed by the
use of an appropriate nozzle which will control the
orientation of ~lattened particles such as sperm heads
or chicken RBC relative to the laser beam of the flow
microfluorometry system (m ; h). As an alternative
approach, distribution artefacts can be tested by
sorting the population into its separate components and
then reanalyzing them independently: if an artefact is
involved, each reanalyzed peak will give a bimodal peak
similar to that observed originally.
Various aspects of the invention are as follows:
A method of sorting living spermatozoa, the method
comprising: the vital staining of ~permatozoa, with a
fluorochrome dye; subjecting the stained spermatozoa to
a light source which causes fluorescence; and sorting
the spermatozoa according to the fluorescence
intensities associated therewith.
A method of sorting living spsrmatozoa, the method
comprising: the vital staining of spermatozoa with a
fluorochrome dye which binds tightly to DNA, does not
intercalate the DNA molecule oP chromosomes and hence
does not disrupt the structure of DNA molecules,
subjecting the spermatozoa to a light source which
causes fluorescence and sorting the spermatozoa into
different groups according to the fluorescence
intensities associated therewith, one group mainly
comprising X-chromosome bearing spermatozoa; and another
group mainly comprising Y-chromosome bearing
sp~rmatozoa.
The dye m y be a bis'oenzimidaæole dye.
In an embodiment of the invention, the bisben-
zimidazole dye Hoechst 33342 is used as a vital
fluoresGent stain for DNA which allows spsrmatozoa to
remain motil~ after analysis. The ~luorescence may be
examined in detail using a commercially
~s~
1 available fluorescence-activated cell sorter.
For a better understanding of tha present
invention, and to show h~w the same may be carried into
effect, reference will now be made, by way of example, to
the accompanying drawings in which:
FIGURE 1 is a graph showing the distribution of
fluorescence of bull spermatozoa stained with Hoechst
33342;
FIGURE 2 is a graph showing the distribution of
Figure 1, with a higher gain setting for the
fluorescence-activated cell sorter;
FIGURE 3 is a graph showing the distribution of
cockerel spermatozoa stained with Hoechst 33342 ~5 ~g/ml)
in egg medium;
FIGURE 4 is a graph showing reanalysis of the
peaks AI and AII in Figure 2;
FIGURES 5a to 5c are graphs showing the results of
analysis with different orientations of the cells; and
FIGURE 6 is a table showing the effect of an
orientating no~zle on FACS analysis of chicken RBC (size~
and bull spermatozoa tfluorescence) compar~d to
non-orientated cells.
In preparation for the analysis semen is
collected, using an appropriate artificial vagina (c),
from Fresian and Hereford bulls. Shortly after
ejaculation, semen is added to 1-2 volumes of egg or milk
medium at 20-22 C. Milk medium is made according to the
method described in (a), which comprises: centrifuging
pasteurized milk at 2000 g for 10 min; removing the
cream; taking the underlying fat-free liquid from this
slow speed spin; and pelleting the milk solids by
centrifugation at 48000 g for 30 mins. The clear
supernatant is then heated at 92-96C for 10 min, and
0.125 g D-fructose/ml and antibiotics (104 units
penicillin ~ 10 mg streptomycin sulphate per 100 ml) is
added when the supernatant has cooled.
The spermatozoa are washed twice by centrifugation
~^
l at lO00 g for 5 min followed by gentle resuspension of
the pellet in sufficient fresh medium to give a
concentration of, for example, 5 x 106 spermatozoa/ml.
Intact spermatozoa are then stained with Hoechst
33342 in milk medium at a concentration of 2 ~g/ml for
bull spermatozoa and 5 ~g/ml for cockerel spermatozoa, at
room temperature for 2-3 hours. The dye concentrations
may be determined empirically from subjective assessment
of optimal staining without overt cytotoxicity.
Flow microfluorometric analysis (g) is carried out
using a fluorescence Activated Cell Sorter (such as, for
example, FACS II:Becton Dickinson Electronics
Laboratories, Sunnyvale, California). The light source
for the FACS may be a 16~-05 ultra violet-enhanced
argon-ion laser, (Spectraphysics), operated at 20 mW in
the u.v. range of wavelengths. Right-angle scatter of
u.v. laser light is prevented from entering the
fluorescence detector by a Wratten 2B filter. The FACS
is calibrated in the u.v. using glutaraldehyde-fixed
chicken red blood cells (f~.
Samples of spermatozoa are analysed and sorted at
room temperature (20-22C) at a rate of up to 3500-5000
cells/sec, except duxing orientation experimen~s in which
the rate was reduced to <800 cells/sec. The sheath fluid
is Dulbecco's phosphate-buffered saline (pH 7.2;
containing Mg2+ and Ca2 ), but without stain.
The total fluorescence is calculated (in arbitrary
units), ~or example by a computer. Such a computer is an
LSI-ll based mini computer (Digital Equipment
Corporation, MA, USA) linked to the FACS, which
calculates the total fluorescence between channels l and
25~ as follows (I):
256no o~ cells in a channel x channel no.
Total fluorescence = ~
t 100 (I)
Cells can be orientated in a single vertical plane
.. ~
v~
1 at a predetermined angle to the ]aser beam by the method
described in (m). A (wedge shaped) sample injection
tube, with faces set at ~0C to the axis flow, has the
effect of making a (central) stream ribbon-shaped within
the sheath stream. Since the velocity of the sheath
stream is considerably higher than that of the sample
stream, the latter is drawn into a thin ribbon and the
flattened cells within this sample become orientated into
the plane of the ribbon.
Extrapolating from maximal flow rates which allow
successful orientation of chicken red blood cells, it has
been estimated, on the basis of cell (head) size and
viscosity of the medium, that successful orientation of
spermatozoa should occur providing that the flow rate
does not exceed 800 cells/sec, when using a sample
density of 5 x 106/mlO
When necessary, heads may be removed -Erom the
spermatozoa in milk medium by ultra~sonication for 5-10
min in a MSE ultrasonicator.
A population of bull spermatozoa stained for a
minimum of 2 hours with Hoechst 33342, (2~g/ml Hoechst
33342) in milk medium shows a complex distribution of
fluorescence intensity, which is illustrated in Figure 1.
Data are given for spermatozoa in milk medium at ambient
temperature (20-23 C) for 2 hours and those killed by
being heated to 56C for 5 min. There are two pairs of
peaks in the distribution, which have been labelled A and
B respectively. When examined microscopically, cells
from window B are non- (or only partly) motile, whereas
spermatozoa sorted from window A show active forward
motility. The likelihood that the B peaks represent dead
or moribund spermatozoa was tested by submitting a sample
of stained spermatozoa to 56 C for 5 min. This treatment
left the spermatozoa totally immotile and when the
fluorescence distribution of these immotile spermatozoa
was examined the entire distribution was concentrated in
the B peaks. A small peak seen between A and B in Figure
--8--
1 1 may represent spermatozoa in a transitory state between
A and B or the presence of a small percentage of diploid
spermatozoa (h).
Attention was concentrated on the A peaks of the
fluorescence distribution of stained bull spermatozoa by
running the FACS fluorescence gain at a higher setting
(Figure 2) so that the B peaks moved off-scale. The low
and high peaks of the observed bimodal fluorescence
distribution of the A peaks (AI and AIII) contained
approximately equal numbers of spermatozoa. The average
fluorescence of spermatozoa in peak AII was approxiamtely
30~ higher than that in peak AI.
Qualitatively similar bimodal distributions are
also obtained using the same procedures as outlined above
for the bull, when analysing ejaculated rabbit, sheep,
goat and human spermatozoa.
When cockerel spermatozoa (~0.5 x 4 ~m heads, ~ 8
~m tails) were stained with H33342 the resulting
fluorescence profile was quite different from that of
bull spermatozoa (Figure 3). The monophasic distribution
of fluorescence may reflect either the homogametic nature
of male birds or be due to the absence of an orientation
artefact in the cylindrically headed cockerel
spermatozoa. The bimodal fluorescence distribution of
bull spermatozoa may be due to a machine artefact,
analogous to that observed for light scatter (size)
analysis of chicken red blood cells, but reflect
underlying biological or physiological differences. An
investigation into the nature of the observed bimodality
was carried out by an analysis-sort-reanalysis of stained
spermatozoa and by the use of an "orientating" nozzle~
First, the living, Hoechst 33342-stained bull
spermatozoa with a fluorescence distribution similar to
that shown in Figure 2, were physically separated
(sorted) into AI and AII population. Each separated
population was then re-analysed and the respective
fluorescence distributions are shown in Figure 4.
,51~ U~3
1 Although the peaks were not clearly unimodal, the
spermatozoa from the AII fraction had a higher overall
fluorescence than those from AI as would be expected if
the spermatozoa in peak AI were from a population
different from that of those in peak AII~ The low
fluorescent peak appearing at approximately channel 30
for both populations in Figure 4 was due to spermatozoa
from which the H33342 had leached. Fixation of
spermatozoa with buffered formal-saline (pH 7.4) before
or after staining or after they had been sorted failed to
reduce the leakage of dye. In 17 experiments in which
the spermato~oa in peaks AI and AII were separated, the
total fluorescence intensity of the reanalysed AII
population was 15.~ ~ 2.9~ greater than that of the AI
population. For a comparison, the same experiment was
performed using chicken RBC. It is known that the
apparent bimodal size distri~ution of the chicken RBC is
an artefact related to the orientation of individual
cells to the laser beam. When the chicken RBC were
sorted into two peaks on the basis of scatter, each
separated peak gave the same bimodal distribution as the
original, unsorted, material when reanalysed.
Second, an orientation nozzle similar to that
decribed in (m) was used to analyse bull spermatozoa.
The efficiency of the nozzle was tested using a
light-scatter analysis of chicken RBC (1200 cells/sec).
Figure 5 shows results using an orientating nozzle for
(a) chicken RBC and (b, c) bull spermatozoaO In Figure
5a) peak 1 was obtained when the sample ribbon was
parallel to the laser beam; peak 2 was obtained when the
sample ribbon was at right angles to the laser beam; and
peak 3 for randomly orientat0d cells. In Figure 5b) peak
1 was obtained when the heads of the spermatozoa were
orientated edge on with respect to the laser beam and
peak 2 when the sample was rotated through 90 in the
axis of the flow (laser beam intersecting the broad side
of head); randomly orientated cells are indicated by 3.
,.~,
~,2~U~
-10~
1 In Figure 5c) the bimodal distribution of fluorescence
intensity of intact Hoechst 33342-stained bull
spermatozoa was not affected by altering the orientation
of the sample ribbon: the distributions of randomly
orientated cells overlapped. The scatter distribution of
chicken RBC (Figure 5a) was affected by orientating the
cells with their edges parallel to or at right angles to
the laser beam. A similar effect was observed when sperm
heads were passed through the orientating nozzle and th
effect on the fluorescence profile examined. Although
bull spermatozoa have flattened heads, they did not
display a biphasic scatter (size) profile similar to that
seen when analysing chicken RBC. Nevertheless, the heads
of bull spermatozoa could be positively orientated, since
the resulting fluorescence profiles were monophasic and
did not overlap; (Figure 5b). In contrast, the bimodal
fluorescence distribution of intact bull spermatozoa
stained with Hoechst 33342 was not altered by rotation of
the nozzle ¦Figure 5c). The percentage of cells within
each peak is shown in Figure 6.
Bull spermatozoa stained with Hoechs-t 33342 in
milk or egg medium show a complex profile of
fluorescence when analysed on the FACS. The observed
fluorescence distribution of particles the size of
spermatozoa (-2 x 5 x 10 ~m head, 40 ~m tail) can be
divided into three main areas: (1) unstained material,
(2) a pair of highly fluorescent peaks ~B) shown to
consist of dead or moribund spermatozoa, and (3) a pair
of peaks (AI and AII) with intermediate fluorescence
which consist of spermatozoa with normal forward
motility. Attention has been concentrated on peaks AI
and AII.
An increased staining of non-viable cells by
~oechst 33342 similar to that seen here for bovine
spermatozoa has previously been reported for dead or
dying lymphocytes stained with the same dye. It has been
suggested (n) that the increased uptake of stain was due
. - ' '
:~,~S~!~U~
1 to a breakdown of the integrity of the cell membrane at
cell deathO This may be the mechanism responsible for
~he observed increase in fluorescence of dead spermatozoa
although it is possible that the normally tightly packed
D~A in the nucleus becomes disorga~ni~ed and this
contributes to the increased staining. However,
preliminary fluorometric studies suggest that a
considerable increase in the fluorescence intensity of
Hoechst 33342 occurs as the pH decreases, irrespective of
whether the dye is bound to DNA, protein or is free in
solution. This observation suggests that the B peaks may
arise because of increased nuclear acidity at death.
The bimodal distribution observed in the Hoechst
33342 staining of viable spermatozoa (peaks A) is
probably a consequence of the biologically different
kinds of spermatozoa in the normal ejaculate.
Accordingly a comparison of the fluorescence profiles of
mammalian and bird spermatozoa, which are heterogametic
and homogametic respectively shows the cockerel
spermatozoa to have a unimodal distribution, Figure 5
illustrates that although the heads of spermatozoa can be
orientated, the bimodal fluorescence distribution of
Hoechst 333~2-stained intact live spermatozoa is
apparently independent of the orientation of the sperm
heads around their long axis; and peaks AI and AII
(Figure ~), althoug~ not clearly unimodal, are of
predictable fluorescence in that spermatozoa separated
from peak AII fluoresce more brightly than those from AI:
a difference which averages at about 15%. If bimodality
had a machine orientation artefact the separated
population would be expected to have identical (bimodal)
distributions.
Thus the observed bimodality of fluorescence
distribution indicates the presence of two
physiologically or biologically different sub-populations
of viable spermatozoa. The sub-populations (AI and AII)
may reflect spermatozoa at distinct stages of late
~l~t~
-12-
1 maturation or the difference between X- and Y- chromosome
bearing spermatozoa. Experimental work with rabbits has
yielded a 3.5:1 ratio of correct sex to incorrect sex,
which is very close to the ratio which would be predicted
from a theoretical estimate of the overlaps between the
two sorted peaks. The above described method thus has a
useful application in sorting spermatozoa according to
whether they are X- or Y- chromosome bearing spermatozoa.
