Fiber DNA studies of premeiotic mouse spermatogenesis

Experimental

Cell Research 146 (1983) 281-287

Fiber DNA Studies GEORGIANA

of Premeiotic

JAGIELLO,’

Mouse

Copyright 0 1983 by Academrc Press, Inc. All rights of reproduction in any form reserved 0014.4827/83/080281-07$02.00/O

Spermatogenesis

WENG KONG SUNG’ and JACK VAN’T HOF2

‘Departments of Obstetrics and Gynecology, Human Genetics and Development and Center for Reproductive Sciences of the International Institutes for the Study of Human Reproduction, College of Physicians and Surgeons of Columbia University, New York, NY 10032 and *Brookhaven National Laboratory, Associated Universities, Inc., Upton, NY 11973, USA

SUMMARY The technique of fiber DNA measurement was used to study the possibility that the lengthening of the DNA “S” phase previously reported for mouse premeiotic spermatogonia was due to a reduced number of initiation sites. The mean replicon size of neonatal mouse preleptotene cells was similar to sizes reported for adult mouse somatic cells. A slow rate of DNA chain growth was observed in all cells from day 1 through days IO-12 of age. It was felt that the meiotic entry in male mouse germ cells may involve a slower replication fork rate and other factors which increased the time between activation and replication of replicon families.

Much attention has been devoted to the duration and character of the DNA synthesis period (“S” period) of somatic cells, but there is little published information about the premeiotic DNA synthesis period of germ line cells. A progressively lengthening “S” period in the cell cycles of murine oogonium and spermatogonium has been documented [l, 2, 31. Sometime ago Callan suggested that the lengthened “S” period of Triturus spermatocytes was due to a decreased number of initiation sites on the DNA [4]; however, his suggestion remains to be tested in the mouse. Since alterations in the length of “S”, whether exogenously induced or spontantously occurring, would theoretically alter the subsequent phenomena of chiasma formation and recombination [5], an investigation of this putative mechanism in mouse seemed appropriate. The first wave of spermatogonial mitoses occurring in the neonatal Swiss mouse was used as a test system for assessing fiber DNA size as an index of the number of initiation sites. MATERIALS

AND METHODS

Five neonatal Swiss CD-l (CAMM) male mice were sacrificed by cervical dislocation on each day of life from day 1 through 12. The testes were quickly removed, washed three times in PBS and placed onto a grid in an organ culture dish (Falcon) containing 1 cc of McCoy’s 5A medium (Gibco) with added 5 mM I-glutamine, penicillin, 100 units, and streptomycin, 100 units, and 10% gelded horse serum (Gibco). The meniscus of the medium reached the midpoint of the outer surface of each testis. Fifty uCi of [3H]thymidine (New England Nuclear, NET-027Z, 50-80 Ciimmol) were injected directly into each testis. Culture was carried out at 32°C the normal in vivo testis temperature, in a 5 % COJair, high humidity atmosphere. Two testes were removed every 30 min up to 2 h and washed three times with PBS. They were decapsulated and incubated with 50 ml of 0.25 % collagenase in PBS at 37°C in a shaking water bath for 30 min. The separated cells were washed with PBS and spun three times at 500 rpm for 5 min to remove interstitial cells. Single testicular tubules were removed from the pellet and incubated with 50 ml of 0.25 % trypsin in PBS at 37°C in a shaking water bath for 30 min. Single cells thus dislodged were then washed three times with PBS and resuspended in 1 ml PBS. These were processed by the method of Van? Hof [6, 71 for obtaining DNA fibers. The method used is a modification of that developed by Lark et al. [8]. Briefly, it involves removal of an aliquot of suspended cells, placement of the cells on a subbed microscope slide, incubation of the cells with about 25 ul of trypsin (I mg/ml) for 15 min at 37”C, the addition of 25 ul of 1% SDS in 19-838333

282 Jagiello,

Sung and Van’t Hof

Exp Cell Res 146(1983)

Fig. 1. DNA fiber from day 10 (a) and day 1 (b) mouse testicular cell (0.5 h); Planapo 10 X bright light lens and a phase III condenser were used. (6) Each segment (m) was measured to calculate the rate. (c) DNA fiber from day 1 mouse testicular cell (2.0 h); whole length of fiber was measured to obtain the replicon size. (6) Day 1 mouse testis; gonocytes ( T ) were not labelled after 2.0 h incubation with [3H]TdR. (e) Histological section of day 10 mouse testis; preleptotene spermatocytes ( t ) were labelled after 0.5 h incubation with [3H]TdR. cf) Cytological preparation of day 11 mouse testicular cells; preleptotene ( t ) and leptotene spermatocytes were labelled after 0.5 h incubation with [3H]TdR. (a-c) x150; (d-f) x1600.

IO-* M EDTA, further incubation for 15 min at 37°C. the addition of 25 ul of lo-* M EDTA, and finally, slightly tilting the slide to allow the mixture to slowly spread downward on the surface of the slide. Once spread, the mixture was air-dried and the slides processed as described by Huberman & Tsai [9]. Autoradiographs were prepared using Kodak NTB liquid emulsion. Slides were exposed for 3 months at 4°C. They were developed in Kodak D-19 at 22°C for 12 min and fixed. Photographs were taken of the fibers developed on the slide with a Zeiss Photomicroscope II using a Planapo 10X bright light lens and a phase III condenser (fig. 1a). Panatomic-X (Kodak) film was used. The photographs were enlarged and printed to a total magnification of x500. The labelled DNA segments of two or more sequentially aligned pairs were then selected, measured as described by Van? Hof 16, 71 with a Manostat micrometer and tabulated. It was assumed that replication within a replicon occurred bidirectionally and that the rate of DNA chain growth by each of the two replication forks within a replicon was obtained by measuring the

DNA studies of mouse spermatogenesis

Exp Cell Res 146 (1983)

Rate (pm/hr)

gi$/k

k

0 102030405060 70

Size

0 10203040506070

Fig. 2. Frequency distribution of rate of DNA chain growth after a pulse of 0.5 h in each age group; cells from testes of neonates aged (A) 1 and 2 days; (B) 3-5 days; (C) 6-9 days; (0) l&l2 days. Fig. 3. Replicon-size determined from DNA fibers of cells labelled for 0.5 h. Fig. 4. Frequency distribution of replicon size in each age group after a pulse of 2.0 h.

(pm)

length of each contiguously labelled segment of sequentially aligned pairs of tandem grain arrays (fig. 1b). The average rate of replication fork movement was calculated by dividing the average length of each labelled segment of each pair by the duration of the isotope pulse [7]. This method provides reliable estimates of single fork rates but only if the duration of the pulse is less than the time needed for the converging forks of neighboring replicons to meet, stop moving, and fuse to produce a single array. The lack of fork movement during the course of the radioactive pulse results in an underestimation of the rate of chain growth. To avoid underestimations the ratio (replicon size)/(pulse duration) x 2 (single fork rate) must be greater than one [lo]. Replicon size was estimated by selecting long individual fibers with tandem arrays of grains, determining the length of the fiber, dividing the length by the sum of the arrays arranged on it, and multiplying by 2 [6, 71 (fig. 1c). Again, estimations obtained by this method are reliable provided the duration of the pulse does not exceed the time needed for adjacent converging forks of neighboring replicons to fuse and form a single array. When the pulse duration exceeds this time limit replicon-size estimations increase resulting in an overestimation [ll]. Two control testes were removed on each day of the experiment to ascertain cell populations and mitoses. Tissue was immediately processed for histologic examination. Cytological preparations by the Meredith method [12] were also made of a small piece of testis from days 9-12 to more accurately differentiate the spermatogonia B, preleptotene and leptotene spermatocytes. The testes for histologic examination were immediately immersed in 4 % glutaraldehyde in sodium cacodylate buffer at pH 7.4 for 15 min. They were then cut into small pieces and kept in the fixative at 4°C for an additional 2 h. They were then dehydrated in a series of ethanol, embedded in Epon 812 and serially sectioned on a Porter Blum Microtome at 2 pm. Slides for autoradiographs of the histologic sections and cytological preparations were treated and developed as above with the exception that the exposure period was 2 weeks. They were then stained with 1% toluidine blue and 5 % Giemsa respectively. Cell populations of the tubules were classified into Sertoli cells by the usual histologic criteria and the stages of spermatogenesis by a modification of the staging of Oakberg [13] and Bell& [14]. One hundred tubules were examined from each pair of control testes. Five hundred nuclei prepared by the Meredith method were scored for meiotic stage for each pair of testes from day 9 through 12 and tabulated as noted above.

283

284 Jagiello,

Sung and Van’t Hof

Exp Cell Res 146 (1983)

RESULTS The control study defined the cell populations which were present at age 1-12 days. The first wave of meiosis was observed in this study, and in a separate study of this strain (Sung & Jagiello, unpublished observations) to occur on day 10 as indicated by a sharp increase in preleptotene spermatocytes. Testes from animals aged 1 and 2 days (group 1) contained 96% Sertoli cells, and 4 % gonocytes where no gonocyte was labelled with [3H]thymidine (fig. 1 d). Between days 3 and 5 (group 2), Sertoli cells remained almost constant at 94%, whereas the remainder of the spermatogenic epithelium was made up of 1% gonocytes, 4.5% primitive type A spermatogonia, and 0.5% spermatogonia A-B. Tubules from ages 6-9 (group 3) days contained 67 % Sertoli cells, 3 % primitive type A spermatogonia, 29.4 % A-B spermatogonia and 0.6 % preleptotenes. Between ages of 10 and 12 days (group 4), the Sertoli cells made up 32 % and the germ cells were composed of 19.5% A and B spermatogonia and 48.5% preleptotene spermatocytes where all types of germ cells including preleptotene were labelled (fig. 1 e, j). Autoradiograms of 3H-labelled DNA fibers isolated from pre-meiotic cells of newborns of the four developmental age groups showed that the replication fork rate, determined from grain arrays produced by a 0.5 h pulse, was nearly identical in all cases. The data when expressed as pm/h produced histograms with modes between 10 and 15 pm/h and a range from 5 to 45 urn/h (fig. 2). Between 67 and 71% of the estimated fork rates were in the 10-20 pm/h range with the lower percentage representing rates of replicons from the youngest and the higher percentage those from the oldest age group. The mean fork rate, however, was similar regardless of the developmental age of the cells (table 1). Replication fork rates determined from autoradiograms of DNA fibers isolated from cells labelled for more than 0.5 h produced values that were less than 16 pm/h. These lower values indicated that the average replicon completed the replication of the DNA allotted to it within 1 h thereby producing an underestimation of the fork rate. Replicon-size measurements from autoradiograms of DNA fibers from cells labelled for 0.5 h are presented as histograms in fig. 3. The distributions show that the replicon size was nearly the same, regardless of the pre-meiotic stage of the Table 1. Rates of DNA chain growth after a pulse of 0.5 and 2.0 h in each group dividing

by the days after birth

Group

No. of segments

1 Oh) 2 @-3

618 671

3 (b-9) 4 Uh~d

656

Grand mean

590

Rate in 0.5 h Mean f SE (wW

No. of segments

Rate in 2 h Mean f SE (w.nW

16.25kO.23 16.44+0.21 17.9OkO.34 16.00+0.26

256 361 495 464

8.64+0.31 9.46kO.23 10.27kO.25 7.34kO.24

16.6420.85

8.92k1.25

DNA studies of mouse spermatogenesis

Exp Cell Res 146 (1983)

cells from which the DNA was isolated. In each instance the mode was between 15 and 20 urn, the sizes ranged from 5 to 55 urn, and the mean size was between 17 and 20 urn (table 2). Given a replicon size of about 18 urn and a single replication fork rate of approx. 16 urn/h, an average single replicon would replicate the length of DNA allotted to it within 0.56 h, assuming bidirectional replication. A 0.5 h pulse is within this time limit but those of 1 h or more exceed it. Consequently, these longer pulses will give larger replicon sizes, most of which are overestimations. This predictable change to larger estimations of replicon size was observed in the present experiments. Autoradiograms of DNA fibers from pre-meiotic cells labelled for 2 h gave replicon-size measurements that ranged from 5 to 70 urn and, in three of the four cases, modes that were between 20 and 25 urn (fig. 4). The wider range and the shift of the modes toward larger sizes on DNA labelled for 2 h as compared with those seen on fibers labelled for 0.5 h are indicative of fusion of converging forks of adjacent replicons. As a result of this phenomenon the mean sizes of the four groups of pre-meiotic cells estimated with a 2 h pulse exceeded those measured with a 0.5 h pulse by a factor of 1.4-1.8 (table 2). An important aspect of the analysis of DNA fiber autoradiograms is to understand how the observed fork rate and replicon size depend on the duration of the isotopic pulse. This dependence is demonstrated in fig. 5 with data obtained from fibers of pre-meiotic cells from tubules of testes of newborns 10-12 days old. Here, the mean fork rate is expressed as a function of pulse time and the curve produced is linear with a negative slope. Extrapolation of the curve to zero pulse time causes it to intercept the ordinate at approx. 18 urn/h, a value slightly more than the 16-17 urn/h obtained with a pulse of 0.5 h (table 1). In fig. 5, the mean replicon size is shown to increase with pulse time and in this case extrapolation of the curve to zero pulse time gives a value of about 18 pm which is 2 urn more than the 16 urn estimation obtained with a 0.5 h pulse (table 2).

That the values of replicon size and fork rate obtained by extrapolation to zero pulse time are within 12% of those measured on fibers labelled for 0.5 h adds support to the contention that the replicons of pre-meiotic cells, particularly those in the pre-leptotene “S” phase, are about 18 urn in size and have replication forks that travel at 16 pm/h.

Table 2. Replicon size in each group after a pulse of 0.5 and 2.0 h

Group

1 KU

No. of fibers 148

2 C’-5)

158

3 (Dm) 4 (DIMZ)

155 160

Grand mean

Size in 0.5 h Mean + SE (pm)

No. of fibers

Size in 2 h Mean 2 SE (pm)

19.57kO.62 18.24kO.52 17.62LO.54 18.31+0.70

81 109 177 136

28.lOk1.48 30.10+1.19 31.64fl.09 27.722 I .32

18.43+0.81

29.39* I .82

285

286 Jagiello,

Sung and Van’t Hof 30-

? 3

28_

i

26 24-

Exp Cell Res 146(1983)

5. Change in (A) replication fork rate; (II) replicon-size as a function of the pulse duration with [3H]thymidine measured on DNA fibers isolated from premeiotic cells in testicular tubules of mice 9-12 days old. Linear regression equation for fork rate: Y= --5.35X+ 18.05, correlation coefficient equation replicon-size: 0.97; for Y=6.44X+ 15.62, correlation coefficient

Fig.

lm

f5 22 ”

20-

34 16- ,/’ F (r

PULSE’(hours)

16~’

,I

0.5 1.0 1.5 PULSE (hours)

I

2.0

0.98.

DISCUSSION The observation of Callan [4] based on work with Triturus vulgaris, that decreased DNA initiation sites was a characteristic of preleptotene “S”, was not found in the present study with mouse spermatogenesis. The mean replicon size of mouse preleptotene cells was determined to be approx. 18 urn, a value that agrees well with the modal size of 20 urn observed in DNA of asynchronously dividing mouse L tibroblasts [ 151and with the 14 urn size reported for mouse 3T3 cells [16]. The striking aspect of the present study was the slow rate of DNA chain growth observed in the entire series of cells. Group 4 contained the greatest number of preleptotene spermatocytes; yet no significant changes were noted to indicate that initiation sites were being inactivated. The data could, however, be interpreted to indicate that the lengthening “S” phase, described by others [2, 31 to be characteristic of mouse spermatogenesis as it approaches the initial meiotic events, is partially a reflection of the rate of DNA fiber growth. The latter is slow compared with mouse libroblasts [ 171which has been reported to be 3ti2 urn/h. On the other hand, the fork rate of mouse 3T3 cells, determined by the bromodeoxyuridine photolysis method, was 12 urn/h for non-transformed cells and about 20 urn/h for those transformed by the virus SV40 [16]. Whether the different fork rates reported for somatic mouse cells is real or merely a reflection of the means by which it is measured remains an open question. In the present experiments we chose not to use 5-fluoro-2’-deoxyuridine to reduce the DNA precursor pool size, because it has been demonstrated by others that this inhibitor effects the pattern and frequency of replication initiation sites [ 15, 181,as well as chain elongation [ 171. Nevertheless, it is possible that the concentration of thymidine we used to label the nascent chains of premeiotic mouse cells may, in some unknown way, have lowered the rate of fork movement. Whereas this possibility cannot be ruled out, there is little or no evidence in the literature to support it. The concentration of thymidine we used was 1.4~ 1O-5 M and this amount is very near to that (6.5~ 10d5 M) determined for TCphage-infected Escherichia cells when the rate of DNA syntheis was maximal [ 191.The slow rate of 16 urn/h is characteristic of testicular cells from day 1 through the meiotic switch of days 10-12. Thus the mechanism of the entry into meiosis of the male mouse genome may involve both a slower replication fork rate and factors responsible for an increased time interval between the activation and replication

Exp Cell Res 146(1983)

DNA studies of mouse spermatogenesis

of replicon families that function sequentially as the cells progress through the “S” phase. These investigations were supported by grants from the NICHD (HD-10037 and HD-05862).

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Crone, M, Levy, E & Peters, H, Exp cell res 39 (1965) 678. Monesi, V, J cell biol 14 (1962) 1. Ghosal, S K & Mukherjee, B B, Can j genet cytol 13 (1971) 672. Callan, H G, Proc r sot Lond B 181 (1972) 19. Henderson, S A, Ann rev genet 4 (1970) 295. Van? Hof, J, Exp cell res 103 (1976) 395. Van? Hof, J & Bjerknes, C A, Chromosoma 64 (1977) 287. Lark, K G, Consigli, R & Toliver, A, J mol biol 58 (1971) 873. Huberman, J A & Tsai, A, J mol biol 75 (1973) 5. Blumenthal, A B, Kriegstein, H J & Hogness, I S, Cold Spring Harbor symp quant bio138 (1974) 205. 11. Van? Hof, J & Bjerknes, C A, Exp cell res 136 (1981) 461. 12. Meredith, R, Chromosoma 26 (1969) 254. 13. Oakberg, E F, Am j anat 99 (1956) 391. 14. Bellve, A R, Cavicchia, J C, Millette, C F, O’Brien, D A, Bhatnagar, M & Dym, M, J cell bio174 (1977) 68. 15. Cohen, J E, Jasny, B R & Tamm, I, J mol biol 128 (1979) 219. 16. Kapp, L N, Park, S D & Cleaver, J E, Exp cell res 123 (1979) 375. 17. Hand, R & Tamm, I, J cell biol 58 (1973) 410. 18. Ockey, C H & Saffhill, S, Exp cell res 103 (1976) 361. 19. Mathews, C K & Sinha, N K, Proc natl acad sci US 79 (1982) 302. Received October 14, 1982 Revised version received February 2, 1983

Printed

in Sweden

287