the y cells as secondary spermatocytes,the 6 cells as earliest spermatids, the K cells as spermatids, and the X cells as mature spermatozoa. Introduction.
Proceedings of the National Academy of Sciences Vol. 65, No. 1, pp. 192-199, January 1970
The Separation, Physical Characterization, and Differentiation Kinetics of Spermatogonial Cells of the Mouse D. M. K. Lam, R. Furrer, and W. R. Bruce DEPARTMENT OF MEDICAL BIOPHYSICS, UNIVERSITY OF TORONTO, CANADA
Communicated by Leon 0. Jacobson, October 27, 1969
Abstract. By means of the velocity sedimentation technique for cell separation, single cell suspensions from the testes of the mouse could be separated into at least seven peaks each with a different sedimentation velocity. These were named and characterized as follows: a (12 mm/hr); # (10 mm/hr); y (6.5 mm/ hr); a (4.3 mm/hr); 0 (broad peak with median 2.5 mm/hr); K (1.1 mm/hr); X (0.75 mm/hr). Tritiated thymidine was injected into thirty groups of mice. The spermatogonial cells of each group were separated at one hour and then at daily intervals, and the acid insoluble activity of each fraction was measured. This method enabled us to determine the differentiation patterns of mouse spermatogenesis by following the thymidine label with time. It was found that the spermatogonia and primary spermatocytes formed the 0 peak with S-phase spermatogonia and primary spermatocytes having a similar sedimentation velocity of 2.1 mm/hr. The ,3 cells were identified as late pachytenes and diplotene cells, the y cells as secondary spermatocytes, the 6 cells as earliest spermatids, the K cells as spermatids, and the X cells as mature spermatozoa.
Introduction. Spermatogenesis provides a unique system for the study of cell differentiation in higher animals. Spermatogonial cells differentiate in a highly synchronous fashion into only one type of differentiated cell, the spermatozoon,1 and in this process the differentiating cells exist in both the diploid and haploid forms. However, the potential of mammalian spermatogenesis to provide an insight into the mechanisms of differentiation has been severely limited by our inability to separate and characterize the various classes of spermatogonial cells. Recently, Miller and Phillips have demonstrated that cells of different sizes may be separated efficiently on the basis of their sedimentation velocities.2 Since spermatogonial cells differ greatly in size, they should be separable by this method. In this paper we describe the use of velocity sedimentation for the separation of mouse spermatogonial cells. In addition, we examine the kinetics of differentiation during spermatogenesis by an approach that combines the separation method and thymidine labeling. Methods. Preparation of cell suspensions: Suspensions of single testis cells were prepared by a method designed to produce a minimum of trauma to the cells. The tubules from the testes of 8- to 10-week-old (C3H X C57BL)F1 mice (View-Farm, Clinton, Tenn.) were placed on a petri dish in a small quantity of phosphate-buffered saline (PBS) 192
VOL. 65, 1970
MEDICAL SCIENCES: LAM, FURRER, AND BRUCE
193
and were cut ifito short segments by an array of 50 stainless steel razor blades with edges spaced at 0.2 mm. The suspension was washed with 5 ml of PBS into a centrifuge tube where it was pil)etted up and down a Pasteur pipette several times and allowed to settle for 5 min. The tubules, then virtually free of cells, sedimented to the bottom free from the desired cell suspension. Separation method: The sedimentation chamber (Fig. 1), similar to that designed by Miller and Phillips, was loaded at a rate of about 10 ml/min with 20 ml PBS solution followed by 10 ml of cell suspension containing 2 X 107 testis cells in 0.5% bovine serum albumin (BSA, Calbiochem, Fraction V, B grade) dissolved in PBS, and 750 ml of a 1-3% linear BSA gradient in PBS. After 4 hr of sedimentation at 40C and unit gravity the sedimentation chamber was drained from the bottom at a rate of about 10 ml/min and fractions were collected with an automatic fraction collector. 135 4 0.5 fractions of 5.5 ml were collected, each fraction corresponding to a sedimentation distance of about 0.5 mm. The numbers of cells in each fraction were counted with a Coulter counter (model F, Coulter Electronics Inc., Hialeah, Fla.) fitted with a 70 j diameter aperture. To facilitate the counting, both the PBS and BSA were filtered prior to their use (0.45 , pore diameter, Millipore Filter (a) Corp., Bedford, Mass.). The counts were checked (b) against hemocytometer measurements. The volume of the cells in each of the fractions was determined with (c) a Coulter counter and pulse height analyzer (Northwestern NS-601, 256 channel) calibrated with particles of known volumes.3 To measure the acid insoluble tritiated thymidine (sp. act. 5 Ci/mM) radioactivity, cells from each fraction were filtered on a Millipore filter (0.45 IA pore diam-( eter), washed with PBS, 5% trichloroacetic acid and ' Ccmd (e) 95% ethanol. The acid insoluble activity on each filter was then measured with a liquid scintillation counter. FIG. 1.-Cross section of the Results. Cell separation by velocity sedimencylindrical chamtation : The results of a typical cell separation by ber the experiments. used in sedimentation a,
velocity sedimentation (Fig. 2) demonstrate that layer of PBS (20 ml); b, layer the testis cells can be separated into at least seven of 2 X 107 cells in 0.5% BSA c, (10 3ml); gradient of 1 classes each with a different sedimentation veloc- to (750 ml); % BSA linear d, from ity. These are named and characterized as fol- linear gradient maker; e, to lows: a (about 12 mm/hr); 83 (10 4 0.5 mm/hr); fraction collector. y (6.5 + 0.2 mm/hr); a (4.3 4 0.1 mm/hr); 0 (broad peak with median 2.5 mm/hr); K (1.1 ± 0.05 mm/hr); X (0.75 i 0.05 mm/hr). The sedimentation velocity of a spherical particle depends linearly on both the difference between the density of the particle and the medium, and the crosssectional area of the particle. Thus the separation observed in Figure 2 may be due to differences between the density or the volume of the particles. In order to determine which of these effects was predominant, a suspension of testis cells was prepared, separated and the volumes of the separated cells were measured. Volumes of cells in each fraction are plotted as a function of sedimentation velocity for that fraction in Figure 3. The graph, which is on logarithmic scales, shows that the cell volume varies as sedimentation velocity to the 1.5 power over a range of cell volumes from 60 to 3000 ;A3. Thus the sedimentation velocity of most of the cells depends linearly on the cross-sectional area. Except for the
MEDICAL SCIENCES: LAM, FURRER, AND BRUCE
194
12
14
Sedimentation velocity( mm/hr) 2 4 6 8 10
PROC. N. A. S.
0
e 54
FIG. 2.-Velocity sedimentation separation of spermatogonial cells of the mouse. The peaks in thelcurves are referred to by the Greek letters.
13
I0
0
20
40
60
80
100.
120
Fraction number
considerably denser X cells, the other cells have the same density. Figure 3 shows the modal volumes for each of the major cell fractions. These are: a (3000 1A3); #3 (2040 M3); -y (1090 IA3); 5 (610 IA3); 0 (240 M3); K (72 p3); X (22 p3). Morphological characterization of the sedimentation classes of spermatogonial cells: The previous results show that the various classes of spermatogonial cells can be separated from each other on the basis of their size differences by velocity sedimentation. It was next of interest to determine whether these classes of cells could be identified morphologically. To this end cells from each of the fractions were examined after staining with Giemsa. It was found that the peak labeled X (Fig. 2) consisted of spermatozoa, the K cells were spermatids in various stages of maturation, and the g cells were composed of diplotene and late pachytene cells. Morphological identification of the other classes of cells was less certain. Kinetic studies of differentiation using thymidine labeling and cell separation: Spermatogenesis in the mouse involves the following steps:' Sperma1st meiotic division mitosis secondary sper> primary spermatocy tes (2ni) togonlia ('211) 2n'l reiotic division
m Spelratids (n) -spermrnatozoa(n). Among the matocytes (11) classes of cells participatinig in this l)process, only spermatogonia and primary
VOL. 65, 1970
MEDICAL SCIENCES: LAM, FURRER, AND BRUCE
19.3
spermatocytes synthesize DNA, the former in preparation for mitosis and the latter for meiosis.5 With the availability of a method for separation of spermatogonial cells, and since differentiation in this system is synchronous, it should therefore be possible to label the DNA of spermatogonia and primary spermatocytes specifically, and to follow the movement of the label through all the stages of differentiation. Slope X 1.50
1000
FIG. 3.-A logarithmic plot of the median volumes of the cell fractions versus their respective sedimentation velocities. Ar-, rows correspond to the peaks of cell number observed in Fig. 2.
I
t a
-i _ q3
100
f
-
/
K
A/-/ 10 I
. .I 10 5 Sedimentation velocity (mm/hr)
As a preliminary to the kinetic experiment, it was first important to determine whether thymidine labeling was restricted to only one or two of the sedimentation classes of cells and to determine which of these were involved. Accordingly, mice were injected with tritiated thymidine and were sacrificed one hour later. A suspension of testis cells was prepared and separated as previously described. Typical results of such an experiment are shown on Figure 4, panel 1. For comparison purposes, sedimentation values for each of the peaks (Fig. 2) are shown above the graph. It is clear that thymidine labeling is restricted to only a few of the fractions, associated mainly with the 0 peak. The sedimentation velocity of the one-hour labeled cells, however, is at the low velocity side of the 0 peak, being
MEDICAL SCIENCES: LAM, FURRER, AND BRUCE
196
9e
a
e9 KX
S
KX
601
60| 3dhr
2.5
60
2.5
e KX
. 7 days
~~~~~~~~~~~~~.59 days 8
40
S
4A ~~~~~~~~~~~~~~~~~~2.6 U~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ .
40 [ 2.1 80
C20
80Iday 60 6040
PRoc. N. A. S.
days.
16 days
[4.4
2.8
18 days . l days 8.4u, a r|;20ds
2days
60
2f
:40
9
.
.
202.6
8020
_X _ _ ;_;J [4 FIG2 3.
[
4.2
II12 days
34days
20 days
40 -
20-6. 22 days0. 11 4.39
Is 2.6 12days
80 60 -
6.53.
2-200
0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
80-2. 2615 days 60 5days
40
velo
50
231 days
U
10 10 40 4 6 0 83010 690 0
so~~~~~~~~~~as dpranaysel 2f 5daysedmnainmto for [eaato
Fhe 4a-Thyngon actiit versus s
n
v
i
h
f
oe
2
lo
r0
In view of these findings, it was then theoretically possible to determine the differentiation pathway in spermatogenesis by initially labeling the DNA of
VOL, 65, 1970
MEDICAL SCIENCES: LAM, FURRER, AND BRUCE
197
spermatogonia and primary spermatocytes with thymidine and then following their differentiation into various classes of spermatogonial cells with time. To this end, twenty-eight groups of three mice each were injected interperitoneally with tritiated thymidine (1 ACi/gm weight of animal, a level known to produce little effect on mouse spermatogenesis6). One to twenty-eight days after injection, groups of these mice were sacrificed, suspensions of testis cells were fractionated, and the acid precipitable activity in each of the fractions was determined. The number of cells within each fraction and the total radioactivity remained constant over this period of time. The activity in each of the fractions observed from 1 hr to 23 days after injection is shown in Figure 4. The thymidine label appears to move from one peak to another in a highly synchronous fashion, and the radioactivity peaks correspond in most cases to the peaks in the cell number which are shown above the figure. On the basis of the kn own steps in differentiation during spermatogenesis, the morphological characterization of some of the cell-size peaks, and the fact that thymidine label initially goes into the spermatogonia and primary spermatocytes, our results can be interpreted as follows: On day 0, only one labeled peak (2.1 mm/hr) is obtained, indicating that the S-phase spermatogonia and primary spermatocytes have similar sizes with a sedimentation velocity of 2.1 mm/hr. One day 1, two peaks of approximately equal magnitude were observed. The peak at 1.5 mm/hr represents the spermatogonial cells that have just undergone mitotic division since these cells are smaller than the S-phase spermatogonia. The peak at 2.5 mm/hr belongs to the primary spermatocytes that have just passed through the meiotic S-phase. On day 2, the peak at 1.5 mm/hr disappears and a single major peak with sedimentation velocities between 2.1 to 2.7 mm/hr is observed. This implies that the day-i labeled 1.5 mm/hr cells have grown in size and either regenerated to become spermatogonia or differentiated into primary spermatocytes. From day 3 to day 6, essentially no change is found in the distribution of thymidine label, indicating that the label remains associated with the same cells throughout this period. On day 7 a dramatic shift occurs, in that half the counts appear in the cells that sediment between 8-9 mm/hr. This indicates that these labeled cells enlarged their volumes sixfold in about one day. From day 9 to day 11, about half of the total counts appear with the f3 cells (10 mm/hr). The gradual disappearance of the previously labeled 0 cells (2.5 mm/hr) can be noticed. On day 12 two-thirds of the counts in the ,B cells disappear and become associated with the y (6.5 mm/hr) and 6 (4.3 mm/hr) cells. It can be seen from Figure 3 that ,3 cells (2040 .3) are approximately two times larger than the y cells (1090 Z3), and four times larger than the 6 cells (610 Z3). The interpretation of our data is therefore that the ,B cells undergo two meiotic divisions in at most 24 hours. From this analysis and from morphological studies, the # cells can thus be identified as diplotene cells, the y as secondary spermatocytes and the 5 as earliest spermatids. Beginning on day 12 and continuing on day 13, activity appears in the K cells
198
MEDICAL SCIENCES: LAM, FURRER, AND BRUCE
PROC. N. A. S.
0~~~~~~~~~~~~~~~r
I
in DNA synthetic phase -5---i-
I
0
00
-~~~~leptotene
S:
zygotene 4I0
-t
_ 3
t
pochytene
"
diplatene
1qI meiotic division
i t
2fd meiotic
division
cytoplasmic toss
cytoplasmic
20
to vas deferens
2 3 l0 9 8 7 6 5 4 Sedimentation velocity (mm/hr)
1 .9.8 .7
FIG. 5.-A schematic representation of the differentiation of spermatogonial cells. The path traced by the cells is in the space given by time and sedimentation velocity. Discontinuities resulting from cell division or cytoplasmic loss are represented by arrows, uncertain pathways by dashed lines.
(1.1 mm/hr). This reduction in sedimentation velocity, and therefore volume, presumably corresponds to the loss of cytoplasm during the maturation of the spermatids. The major changes from day 14 to day 19 involve movement of the label into the K peak, accompanied by the disappearance of the y and later the 5 peak. On day 19, some activity moves from the K peak to cells with velocity between 3.6 to 3.9 mm/hr. Activity remains in this peak from day 20 to day 22 and perhaps returns to the K peak on day 22. Thymidine label first appears on the X peak (spermatozoa, 0.75 mm/hr) on day 22. The first labeled spermatozoa are found in the vas deferens on day 26. Discussion. A summary of our findings concerning the differentiation pathway for spermatogonial cells of the mouse is presented in Figure 5 and the different classes of cells are characterized in Table 1. The combination of DNA labeling and cell separation methods has allowed us to follow the kinetics of differentiation in spermatogenesis. At least seven distinct size classes of cells
VOL. 65, 1970
MEDICAL SCIENCES: LAM, FURRER, AND BRUCE
199
can be distinguished clearly by the velocity sedimentation method. We find, however, that the sedimentation velocities of the day 0 and day 1 labeled cells do not fit those size classes exactly. Since the sedimentation values are extremely reproducible, this indicates that there are probably subclasses of some of the cell populations. These could involve small changes in size or density. The most important differences are associated with cells around the region of the 0 peak TABLE 1. Physical characterization of spermatogonial cells of the mouse. Cell type
03 02 04
Sedimentation velocity (mm/hr)
Volume
(M') 300 240
,Y
2.1 2.5 1.5 10.0 6.5
116 2040 1090
a K
4.3 1.1
610 72
01 K
3.8 1.1 0.7
470 72 22
,8
Time to 3H TDR labeling (days)
0 1 1
7-9 11 12
13 20 21? 22
(2.5 mm/hr). On the basis of results from day 0 and day 1 labeling experiments, we assume that there are a number of subclasses of 0 cells. Similar arguments apply to the K cells. The differentiation pathway of spermatogenesis of the mouse determined by velocity sedimentation and thymidine labeling is in good agreement with previous work using autoradiographic and morphological techniques.7 While our experience with velocity sedimentation separation of spermatogonial cells is only preliminary, it seems probable that the method will greatly facilitate the study of mammalian spermatogenesis. Such a study may in turn prove to be an excellent model system for studies on the mechanisms of differentiation and genetic processes in meiosis in higher organisms. 1 Oakberg, E. F., Nature, 180, 1137, 1497 (1957). 2Miller, R., and R. Phillips, J. Cell. Physiol., 73, 191 (1969). 3 Reed, R. D., T. J. Hughes, W. B. Taylor, and W. R. Bruce, Exptl. Cell Res., 56,435 (1969). 4Rugh, R., The Mouse (Minneapolis, Minn.: Burgess Publishing Co., 1968). 5 In addition, the synthesis of a small amount of DNA during meiotic prophase in the anthers of lilium has been reported: Hotta, Y., L. Parchman, and H. Stern, these PROCEEDINGS, 60, 575 (1968). 6Johnson, H. A., and E. Cronkite, Rad. Res., 11, 825 (1959). 7 Roosen-Runge, E. C., Biol. Rev., 37, 343 (1962).
