Interactions of nuclear proteins with DNA, during sperm differentiation in the ram

(1)

HAL Id: hal-01600837

https://hal.archives-ouvertes.fr/hal-01600837

Submitted on 2 Jun 2020

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Distributed under a Creative Commons Attribution - ShareAlike| 4.0 International

License

Interactions of nuclear proteins with DNA, during

sperm differentiation in the ram

Maurice Loir, D. Bouvier, M Fornells, M. Lanneau, J.A. Subirana

To cite this version:

(2)

9 Springer-Verlag 1985

Interactions of nuclear proteins with DNA,

during sperm differentiation in the ram

Maurice Loir 1' 4, 5, Dominique Bouvier 2, Merc~ Fornells 3, Maryvonne Lanneau 1, and Juan A. Subirana 3

1 INRA, Station de Physiologie de la Reproduction, F-37380 NouziUy, France; 2 Pathologie Cellulaire,

Institut Biomedical des Cordeliers, F-75270 Paris, France; a Unidad de Quimica Macromolecular del C.S.I.C. Escucla de Ingenieros Industriales Diagonal 999, Barcelona, Spain; 4 Present address: INRA, Physiologie des Poissons, Campus de Beaulieu,

F-35042 Rennes Cedex, France; s addressee for reprint requests

Abstract. R a m spermatid nuclei and caput epididymal

sperm nuclei were prepared and treated with D T T under conditions avoiding proteolysis. Whole-mount preparations for the electron microscope were made in the presence or absence of the detergent Joy. The chromatin of the less mature, non-round spermatid nuclei displayed a nucleoso- real organization that gradually disappeared at the time the histones leave the nuclei (elongating spermatids). Diges- tion with micrococcal nuclease suggests that polynucleo- some arrays are scarcer and more accessible to nuclease in the elongating than in the round nuclei, with increasing amounts of D N A becoming devoid of nuleosomes. In the protamine-containing nuclei (elongated spermatids), only smooth filaments were observed, which formed thick fibers by parallel aggregation. The change from a nucleosomal organization to bundles of smooth filaments appeared to result from a complex process involving the transitory presence of conspicuous " k n o b b y fibers" that suggest a period- icity in the organization of the spermatidal proteins along the D N A molecules. X-ray diffraction patterns obtained with protamine-containing spermatid nuclei and with sperm nuclei confirm that the D N A is arranged in smoothly bent bundles of parallel molecules. No higher-order reflections that might correspond to nucleosome structures were detected in the 30-200 A region.

Introduction

While extensive studies have dealt with DNA-protein interactions in the nuclei of somatic cells, little is known about the structure of chromatin in sperm, especially mammalian sperm. This is undoubtedly due to the characteristics of the sperm nucleus, which is tightly condensed, stabilized by disulphide bonds, and resistant to various chemical and physical agents. Electron microscopy and biochemical in- vestigations require decondensation of the sperm chromatin, which is usually carried out by disulfide bond reduction, associated with either proteolysis (Marushige and Marus- hige 1978; Young 1979; Zirkin et at. 1980), a detergent (Calvin and Bedford 1971; Young and Sweeney 1979), or urea and salt (Tsanev and Avramova 1981). Under these conditions the DNA-protein interactions are disturbed, and this has led to contradictory results. A nucleosomal organization has been claimed for mammalian sperm chromatin by Wagner and Yun (1981), Wagner et al. (1978), and

Gusse and Chevaillier (1980). Such a beaded organization, however, was not observed by Lung (1968) and Evenson et al. (1978), whereas Tsanev and Avramova (1981) observed both beaded and smooth fibers. In fact, the most suitable techniques to elucidate chromatin organization in mammalian sperm are those that do not result in nuclear decondensation. Studies of intact nuclei by X-ray diffraction satisfy this requirement. Further, the nucleus of the oldest spermatids is already very similar to that of mature spermatozoa, and the structural organization of chromatin in the oldest spermatids and in the caput epididymal spermatozoa probably do not differ markedly from that in the mature spermatozoa. Because the chromatin of these imma- ture nuclei can be decondensed by disulfide bond reduction without proteolysis or detergent treatment (Loir and Lan- neau 1984), one may expect to observe the chromatin structure, if not in its native state, then at least not very different from its in situ state in these nuclei and in mature spermatozoa.

During spermiogenesis, at the time the histones leave the chromatin and are replaced by the protamine(s), spermatidal proteins bind transiently to D N A (Fig. 1) (Platz et al. 1975; Grimes et al. 1977; Loir and Lanneau 1978; Mayer et al. 1981). While in the round spermatid nuclei the nucleohistones are likely organized as in somatic nuclei (Loir and Courtens 1979), it is not yet known how the D N A interacts with the spermatidal proteins in non-round nuclei. Two types of chromatin have been observed in the non-round spermatids in the mouse (Kierszenbaum and

(3)

305 Tres 1975): a beaded chromatin and a smooth chromatin.

However the occurrence of these two types could not be correlated with changes in the nuclear proteins since these changes were unknown in the mouse until recently (Mayer et al. 1981).

In the present paper we investigated the changes in chromatin organization in ram spermatids using spread chromatin preparations and digestions with micrococcal nuclease. We also looked at the organization of the protamine- containing spermatid and sperm nuclei using X-ray diffraction.

Materials and methods

Testes of sexually mature Ile-de-France rams were used for the study, and these were taken mainly during the breeding season.

Preparation of nuclei for spread chromatin preparations and for digestion with nuclease. Non-round spermatid nuclei (maturation steps 9-15, Clermont and Leblond 1955), which are ethylenediaminetetraacetate (EDTA)-resistant, were prepared routinely from 10-20 g of testis as previously described (Loir and Lanneau 1978). Three m M phenyl- methylsulfonyl fluoride (PMSF) was present in all the solu- tions. Either 10 m M iodoacetamide (IAMD) was present in the homogenization solution (control nuclei) or 3 m M dithiothreitol (DTT) was present throughout nuclear preparation (reduced nuclei).

Round spermatid nuclei (maturation steps 1-8), elongating spermatid nuclei (steps 9-12), and elongated spermatid nuclei (steps 13-15) were prepared by centrifugal elutriation of testicular cells, then by treatment with Triton X-100 according to Loir and Lanneau (1982).

Caput epididymal sperm nuclei were obtained by minc- ing one caput epididymis with razor blades in 10 ml buffer A (10 m M Tris/HC1 p H 7.4, 2 mM PMSF) at 2 ~ C. The suspension was sonicated for 30 s, then centrifuged (1000 g, 10 min) and the nuclei were washed twice in buffer A.

The distribution of the spermatid nuclei at various maturation steps (Clermont and Leblond 1955) in the nuclear preparations was determined by differential counts carried out with a hemocytometer using a phase-contrast microscope.

Preparation of nuclei for X-ray diffraction studies. Nuclei at steps 13-15 were prepared by centrifugal elutriation and treatment with Triton X-100 and resuspended in 10 ml buffer A. They were sonicated briefly (Loir and Lanneau 1978), always for less than 50 s, then purified by centrifuga- tion (1000 g, 30 min) through 1.5 M sucrose in buffer A. They were washed twice in the same buffer then either suspended in distilled water and lyophilized or suspended in 50% ethanol and stored at --17 ~ C.

To prepare ejaculated sperm nuclei, ejaculated spermatozoa were washed twice in buffer A. They were then sonicated or mixed three times for 5 min at high speed in an Ultraturax mixer. In both cases the nuclei were purified and treated with Triton X-100 and sodium deoxycholate according to Sauti6re et al. (1984) before being lyophilized.

Decondensing treatments and spreading of the nuclei.

Histones and acid-soluble spermatidal proteins were ex-

tracted before spreading by two extractions with 0.2 M t-I2SO ~ (Loir and Lanneau 1978).

EDTA-resistant nuclei were incubated at 20 ~ C in solution B [50 m M sodium borate buffer, pH 9.0, 1 mM PMSF, 0.05% (w/v) lima bean trypsin inhibitor (LBTI) plus 5 m M D T T for the reduced nuclei. After 2 h of incubation one volume of 14 mM I A M D in buffer A was added to the two suspensions which were centrifuged (1500 g, 20 rain) each in two tubes. One pellet of each sample was treated to extract the total nuclear proteins, which were analyzed by electrophoresis (Loir and Lanneau 1978). The second pellet was resuspended in 7 m M I A M D in buffer A. Elon- gated spermatid nuclei (steps 13-15) and caput epididymal sperm nuclei were incubated for 2 8 h as above with 5 mM D T T in solution B. Nuclear decondensation was stopped by addition of one volume of 15 m M IAMD. After centrifu- gation, the nuclei were resuspended in 7 m M I A M D in buffer A.

Whole-mount preparations for the electron microscope were made according to Miller and Bakken (1972). The nuclei samples were centrifuged and diluted with either 0.2 m M E D T A or with a 0.3% solution of the detergent Joy (Proctor and Gamble) prepared in water, p H 8. Sam- ples diluted with E D T A were centrifuged (1000 g, 20 rain) 30 min after dilution through a solution of 0.1 M sucrose and 10% formaldehyde onto glow-discharged, carbon- coated grids. Samples diluted with Joy were similarly centrifuged either immediately or 5 rain after dilution. The grids were air-dried after treatment with 0.4% K o d a k Photo-Flo and rotary shadow-cast with platinum. All observations were made with a Philips EM300 electron microscope.

Digestion with micrococcal nuclease. Nuclei from round and elongating spermatid populations prepared by centrifugal elutriation were resuspended in 1 ml 20 m M Tris/HC1, pH 7.5, 0.1 m M CaC12 (12-15 OD25 o per ml) and preheated for 3 rain at 37 ~ C. Digestions were carried out exactly as described by Bouvier et al. (1985) for 1.5 and 3 rain for each nuclear type.

The D N A fragments generated by micrococcal nuclease were deproteinized, precipitated and analyzed in 2% agarose gels according to Bouvier et al. (1985).

X-ray diffraction. Purified nuclei were suspended in distilled water and were centrifuged at 3000 g for 15 ran. The wet pellet was placed in a glass capillary and sealed. In some experiments a drop of saturated NaC1 solution (containing solid NaCI) was also placed in the capillary to dehydrate the sample partially (76% relative humidity). X-ray diffraction patterns of the sealed pellets were obtained in a modi- fied Warhus camera attached to an Elliott GX6 generator as described by Azorin et al. (1980).

Results

Effects of detergent and of DTT treatment on the spreading of the spermatid nuclei

(4)

Fig. 2a. Well-dispersed putative step 9 nucleus. The chromatin has a typical nucleosomal organization. Control nucleus, no Joy. b Marginal chromatin of a step 11-12 nucleus. Transition from the degranulated chromatin released by the anterior part of the nucleus to the beaded chromatin present in the posterior part. Control nucleus, no Joy. c Step 10-11 nucleus. Partly degranulated chromatin, Control nucleus, spread with Joy. Bars represent 0.5 gm

When EDTA-resistant nuclei untreated with DTT (control nuclei) were spread in the absence of the detergent Joy, only the less mature nuclei (step 9 to about step 11) were enlarged and exhibited a wide halo of well-spread chromatin. The step 12 nuclei enlarged only slightly and

(5)

307

Fig. 3a, b. Knobby fibers. Almost all the granules of the knobs are related to smooth filaments which sometimes connect two successive granules (arrows). Control nucleus, no Joy. e Relation between knobby fibers (arrows) and thick branched chromatin threads. Control nucleus, no Joy. Bars represent 0.5 !~tm

nuclei were decondensed to a variable extent a n d roughly retained their shape. After spreading in the presence of Joy, only some E D T A - r e s i s t a n t nuclei estimated to be at step 15 retained their approximate shape a n d were lightly decondensed.

(6)

(7)

9-11), the chromatin consisted solely of beaded filaments (Fig. 2 a). Such beaded chromatin was also observed in the posterior part of the step 11-12 nuclei, whereas only smooth filaments were released from the anterior part of these nuclei (Fig. 2b) or from the posterior part of the oldest step 12 nuclei. Only smooth chromatin filaments were observed in the step 13 15 nuclei (EDTA-resistant nuclei and nuclei prepared by elutriation plus Triton X-100) after 8-h reducing treatments (Fig. 4c).

Changes in chromatin organization

As deduced from measurements carried out on platinum- shadowed preparations, the chromatin of elongatin~ spermatid nuclei showed nucleosome-sized beads (139 A + I 5 , n = 101) with a typical nucleofilament organization (Fig. 2a). When EDTA-resistant nuclei were previously extracted twice with 0.2 M H2SO4, no beaded organization was observed and the chromatin of all the nuclei consisted exclusively of smooth filaments. In step 11 12 nuclei that had been treated or untreated with D T T and spread in the presence of Joy smooth fibers were mixed with fibers having a conspicuously low frequency of beads separated by unequal stretches of smooth filaments (Fig. 2 co). Ih these nuclei the mean diameter of the beads (135 + 14 A, n = 112) was not significantly different from that of the beads observed in the control nuclei spread without Joy.

In degranulated chromatin, in step 10-12 nuclei (control nuclei, no Joy) conspicuous structures were observed. These had the appearance of knobby fibers (Figs. 3a, b) that sometimes originate from coarsely granular chromatin. For chromatin that was not part of these " k n o b s , " the diameter of the fiber was 96 + 12 A (n = 43). The " k n o b s " themselves appeared to be composed of two spherical granules, each with a diameter of 224 + 29 A (n = 73). They were regularly spaced along the fibers, every 1030+28 A (n=42). Usually one or two smooth filaments converged upon these granules. Sometimes, one of these filaments made a short loop usually between the two opposite granules of two successive " k n o b s . " Later, the knobby fibers participate in the formation of branched chromatin threads up to 370 ]~ wide (Fig. 3c).

The anterior part of the step 11-12 (Fig. 2b) nuclei and the late step 12 nuclei released exclusively smooth filaments. The thinner ones were 47+11 ~ wide ( n = 108, measurements carried out on nuclei spread in absence of Joy). In nuclei reduced and spread with Joy in which the smooth filaments first became predominant (Fig. 4a), then the only filaments seen (Fig. 4b), the filaments often coalesced into bundles by parallel aggregation giving rise to typical thick smooth fibers (about 150-380 ~ wide). In more mature nuclei (stepS o14-15) such fibers bearing globular structures 350-1000 A in diameter became numerous and interconnected and gave to the dispersed chromatin a fibro-granular appearance (Fig. 4b) which gave rise to smooth filaments following the most efficient reducing treatments (8 h) (Fig. 4c). The thickness of the thinnest smooth filaments observed in these nuclei was 48 + 12 A (n = 116).

Spread preparations of caput epididymal sperm

When D T T treatment had been long enough to markedly decondense the nuclei, spreading with Joy caused the decondensed chromatin to appear as a network of interconnected fibers and filaments (not shown) of variable thickness ranging between 38-55 A_ and 400 A in diameter.

309

F i g . 5. Agarose gel electrophoresis of DNA from micrococcal nuc-

lease(MNase)-digested spermatid nuclei: Lane A 0 ~174 RF DNA

- Hinc II digest (molecular weight standard), lane B MNase-sensi-

tive chromatin fraction from round spermatids after 1.5 rain of digestion, lane C same as lane B but after 3-min digestion, lane D MNase-sensitive chromatin fraction from elongating spermatids after 1.5 min of digestion, lane E same as lane D, but after 3-min digestion, lane F MNase-sensitive chromatin fraction from HeLa cell nuclei after 3 min of digestion

Digestions by micrococcal nuclease

Figure 5 illustrates the electrophoretic patterns of the D N A extracted from the nuclease-sensitive chromatin fractions of round and elongating spermatids. The pattern obtained with round spermatid chromatin was typical of the enzyme and showed a series of D N A fragments the sizes of which were multiples of a repeating unit estimated to be 170_+4 base pairs (bp) (lane B). The intensity and number of these multiples decreased with increasing times of nuclease action (lane C). Conversely, the D N A extracted from elongating spermatids always displayed the same pattern independent of the duration of digestion. This pattern comprised only two peaks of D N A fragments with sizes estimated at 180_+ 8 and 300_+5 bp (lanes D and E). In addition, a smear of shorter D N A fragments ranging from 50-110 bp was observed in both cell types. They were totally absent from HeLa cell chromatin (lane F) digested under identical conditions. In round spermatid nuclei, the number of such fragments increased with the duration of nuclease action. In elongating spermatids, the size of the smear remained con- stant whatever the duration of digestion.

X-ray diffraction

(8)

Fig. 6. X-ray diffraction patterns obtained from ram sperm nuclei at 100% (a, c) and 76% (b) relative humidities. The outer region of the patterns shown in a and b corresponds to the 3.4 A spacing typical of the B form of DNA. The pattern shown in c was taken by a high-resolution camera. Only a ring with an equivalent Bragg spacing of 26 A can be detected; no higher-order rings are visible within this ring. The patterns shown are essentially identical to those obtained from advanced ram spermatids and from bull spermatozoa

show any feature in the center of the pattern, ind!cating that no higher-order structures of between 30-200 A were present in the sample studied. Nucleosomes should give several rings in this region. The resolution was limited to about 200 ~ by the central blackening around the beam- stop. These observations also argue against a lamellar structure of chromatin in sperm, which has been suggested by Koehler (1966); (3) The patterns also contained a ring at about 12.5-13 A with a shoulder at about 9 A,. This feature indicates that the D N A present in the samples is in the B form (Llopis and Subirana 1975). The appearance of the pattern in this region was similar to the pattern obtained from nucleohistone. The diffuse appearance of these rings indicates that the D N A molecules, although parallel, are randomly arranged in the vertical direction. This observation contrasts with the situation in other protamine-DNA complexes (Suau and Subirana 1977) in which the D N A molecules are organized in a partially crystalline arrangement giving rise to sharp rings in this region of the pattern. (4) A broad peak at 3.5 A was also present, which was the superposition of the 3.4 A reflection due to the base pair stacking in D N A and a contribution of protein.

U p o n moderate dehydration the D N A lost its secondary structure as shown by the fact that the ring at 12.5 A was substituted by a protein ring at about 10 ]t. The behaviour of mammalian sperm is therefore similar in this respect to that of nucleohistone (Llopis and Subirana 1975) and differs from conventional nucleoprotamine complexes (Suau and Subirana 1977), in which the B conformation of D N A is stabilized upon dehydration. These observations indicate that the D N A of mammalian sperm nuclei is organized in bundles of parallel D N A molecules randomly dis- placed in the vertical direction. These bundles may be bent in a smooth manner, but no regularity in these bends can be discerned from the high-resolution diffraction patterns obtained.

Discussion

Spreading of the non-round spermatid chromatin

Because the stabilization of non-round spermatid nuclei of the ram varies markedly throughout maturation (Loir and Lanneau 1984) and because non-covalent and disulfide bonds participate in this nuclear stabilization, it was of

interest to apply Miller and Bakken's (1972) technique to untreated nuclei or nuclei treated with D T T and spread with or without Joy. It has been shown that this method of preparing ram EDTA-resistant nuclei avoids proteolysis (Loir and Lanneau 1978). Spreading of untreated nuclei in the absence of Joy is presumed to induce only minor changes in the native chromatin organization since the nuclei were not exposed to conditions known to dissociate the various spermatid nuclear proteins from DNA. The detergent Joy disrupts non-covalent bonds and so provides a better dispersion of the spermatid chromatin. However, it has been suggested that this detergent may extract some nuclear proteins (Scheer 1978). This effect was not apparent in our material since the presence and size of the nucleosomes as well as the thickness of the smooth filaments were not affected by the use of Joy. Treatment of nuclei with DTT, in the absence of proteolysis, improved the spreading of the elongating spermatid nuclei and allows decondensation of the protamine-containing spermatid and caput epididymal sperm nuclei without significant protein loss. How- ever, it must be kept in mind that, in the observed nuclei, not all the disulfide bonds were cleaved. When they were all reduced, especially in the protamine-containing nuclei (after a 15-h D T T treatment, Loir and Lanneau 1984), the chromatin was so decondensed that the nuclei were either dissolved or failed to sediment onto the grids. For nuclei treated with DTT, spreading with Joy allows the best visual- ization of the chromatin organization in the protamine- containing spermatid nuclei, which are strongly stabilized by disulfide bonds and non-covalent bonds (Loir and Lan- neau 1984).

Chromatin organization in nuclei containing histones and spermatidal proteins

The size of the beads that appeared to be strung out on the chromatin filaments in the step 9-12 nuclei agreed with that observed for shadowed nucleosomes (Oudet et al. 1975). After two extractions with 0.2 M HzSO 4, which re- moved all the histones from the EDTA-resistant nuclei (Loir and Lanneau 1978), beads were no longer observed. These data and the results of nuclease digestions suggest that the beads correspond to nucleosomes.

(9)

311 than from the posterior part. This observation agrees with

histochemical data (Loir 1972; Courtens and Loir 1975) that show that there is an antero-posterior gradient in the removal of the lysine-rich proteins (mainly the histones) from the step 11-12 nuclei.

During micrococcal nuclease digestion, round spermatid nuclei behave as somatic (for instance Hela cell) nuclei, except for a somewhat shorter repeating unit (170_+4 bp in round spermatids versus 180_+26 bp in Hela cells) and for a smear of very short, probably randomly cut D N A fragments. The same average repeat length has been mea- sured in the round spermatids of the rat (172_ 8 bp, Brock et al. 1981) and a band of D N A fragments smaller than 72 bp was observed in micrococcal nuclease digests of decondensed sperm chromatin of the rabbit (Young and Sweeney 1979). Our data therefore suggest that (1) polynuc- leosome arrays are scarcer and far more accessible to nuclease in elongating than in round spermatids so that limited digests are obtained with short nuclease treatments in the former spermatid type, and (2) increasing amounts of D N A become devoid of nucleosomes and momentarily more accessible to the enzyme as maturation proceeds. This might explain the smear that characterizes spermatid nuclei.

The change from a nucleosomal organization to bundles of smooth filaments raises the question of the mechanisms sustaining this transformation. Our present observations al- low us to eliminate the simplest hypothesis of a progressive parallel association between filaments. Indeed, the occurrence of knobby fibers suggests a more elaborate mecha- nism, which should not work along the whole length of the filaments but which should develop at regularly spaced intervals. The knobs might represent structures where the filaments are winding around each other, gradually involving more and more filaments from the interknob portions of the fiber. The diameter of the fiber between the knobs suggests only three parallel filaments. Later, the knobby fibers might come together to build up branched threads in which the knobby organization is no longer visible. The conspicuously regular arrangement of the knobs likely re- flects some periodic protein pattern along the D N A molecules.

Chromatin organization in protamine-containing nuclei

After being spread without Joy, all EDTA-resistant nuclei that were identifiable as being at the end of step 12, had only smooth fibers at their posterior margin. When prepared by elutriation followed by Triton X-100, all the step 13-15 nuclei, treated with DTT and spread with Joy, exhibited only smooth filaments. This was also true for the caput epididymal sperm nuclei. Great care was taken to avoid proteolysis during the preparation of the nuclei, especially during the DTT treatment. Electrophoretic analysis con- firmed the similarity of the basic protein contents of the reduced and unreduced nuclei. For these reasons, we reject the hypothesis that the absence of a beaded chromatin organization in the protamine-containing nuclei resulted from experimental conditions. Using different experimental conditions, Tsanev and Avramova (1981) have decondensed ejaculated sperm nuclei of ram and observed chromatin fibers bearing beads of approximately nucleosomal size. Be- cause reduction alone does not induce the decondensation of mature sperm nuclei we do not have ultrastructural data concerning the organization of the chromatin of these nuclei.

The average diameter of the thinnest filaments observed in steps 13-15 nuclei was 48 A, i.e., significantly greater than that of the D N A double helix (15-25 A, Oudet et al. 1975). However, this value agrees with values obtained on shadowed samples for D N A covered with proteins (McMaster-Kaye and Kaye 1980; Scheer 1978). In this study, the diameter of 48 A corresponds to the DNA-protamine complex. The width of the smooth filaments present in the step 11-12 nuclei was 47 ~. In these nuclei the D N A was covered by spermatidal proteins as well as by some protamine molecules. The presence of spermatidal proteins did not appear to change significantly the diameter of the DNA-protein complex.

The proposal of the non-nucleosomal organization of the protamine-containing nuclei was unambiguously sus- tained by the X-ray diffraction data. The typical rings pro- duced by nucleosomes (Azorin et al. 1980) that occur at about 10, 55, and 37 A were not found in wet nucleoprotamine (Fig. 6 c).

Therefore, our results show that the chromatin of ram protamine-containing spermatids and of the sperm does not possess a nucleosomal organization. Such a conclusion agrees with the experimental data obtained by Kierszen- baum and Tres (1975) in the mouse, Young and Sweeney (1979) in the rabbit, as well as with the proposal of Pogany et al. (1981) who, after theoretical considerations, have con- cluded that in mouse sperm nuclei the D N A cannot be packaged in nucleosomes.

Acknowledgements. The authors gratefully acknowledge Dr. R.J. Kilgour for his help in the preparation of the English text. They also thank M. Terriot for excellent assistance in preparing figures.

References

Azorin F, Martinez AB, Subirana JA (1980) Organization of nucleosomes and spacer DNA in chromatin fibres. Int J Biol Macro- tool 2:81-92

Bouvier D, Hubert I, Feve AP, Bouteille M (1985) Characterization of lamina-bound chromatin in the nuclear shell isolated from Hela cells. Exp Cell Res 156: 500-512

Brock WA, Williams M, Tien Kuo M, Meistrich ML (1981) Sensi- tivity of rat testis nuclei to nuclease digestion. J Cell Biol 91:74 a Calvin HI, Bedford JM (1971) Formation of disulphide bonds in the nucleus and accessory structures of mammalian spermatozoa during maturation in the epididymis. J Reprod Fertil Suppl 13 : 65-75

Clermont Y, Leblond CP (1955) Spermiogenesis of man, monkey, ram and other mammals as shown by the "periodic acid- Schiff" technique. Am J Anat 96 : 229-253

Courtens JL, Loir M (1975) Mise en 6vidence par cytochimie ul- trastructurale de la migration des histones riches eu lysine au cours de la spermiogen~se du B~lier. J Microsc Biol Cell 24: 249-258

Evenson DP, Witkin SS, de Harven E, Bendich A (1978) Ultra- structure of partially decondensed human spermatozoal chromatin. J Ultrastruct Res 63:178-187

Grimes SR, Meistrich ML, Platz RD, Hnilica LS (1977) Nuclear proteins transitions in rat testis spermatids. Exp Cell Res 110:31-39

Gusse M, Chevaillier P (1980) Electron microscope evidence for the presence of globular structures in different sperm chroma- tins. J Cell Biol 87:280-284

Kierszenbaum AL, Tres LL (1975) Structural and transcriptional features of the mouse spermatid genome. J Cell Bio165:258-270 Koehler JK (1966) Fine structure observations in frozen-etched

(10)

Llopis R, Subirana JA (1975) X-ray diffraction studies of calf thy- mus nucleohistone. Ann Quim 71 : 898-906

Loir M (1972) M6tabolisme de l'acide ribonucl6"ique et des pro- t6ines dans les spermatocytes et les spermatides du Bblier (Ovis aries). III. Etude cytochimique des transformations prot6iques dans les vieilles spermatides. Ann Biol Anim Biochim Biophys 12:531-544

Loir M, Courtens JL (1979) Nuclear reorganization in ram spermatids. J Ultrastruct Res 67:309-324

Loir M, Lanneau M (1974) Separation of ram spermatids by sedi- mentation at unit gravity. Exp Cell Res 83:319 327

Loir M, Lanneau M (1978) Transformation of ram spermatid chromatin. Exp Cell Res 115 : 231-243

Loir M, Lanneau M (1982) A strategy for an improved separation of mammalian spermatids. Gamete Res 6:179 188

Loir M, Lanneau M (1984) Structural function of the basic nuclear proteins in ram spermatids. J Ultrastruct Res 86:262-276 Loir M, Bouvier D, Fornells M, Lanneau M, Subirana JA (1983)

Transformation and stabilization of ehromatin in ram spermatids. In: Andr6 J (ed) The sperm cell. Martinus Nijhoff, The Hague, pp 218-221

Lung B (1968) Whole-mount electron microscopy of chromatin and membranes in bul[ and human sperma heads. J Ultrastruct Res 22:485-493

Marushige Y, Marushige K (1978) Dispersion of mammalian sperm chromatin during fertilization: An in vitro study. Bio- chim Biophys Acta 519:1-22

Mayer JF, Chang TSK, Zirkin BR (1981) Spermatogenesis in the mouse. 2. Amino acid incorporation into basic nucleoproteins of mouse spermatids and spermatozoa. Biol Reprod 25:1041-1051

McMaster-Kaye R, Kaye JS (1980) Organization ofchromatin during spermiogenesis: Beaded fibers, partly beaded fibers, and loss of nucleosomal structure. Chromosoma 77:41-56 Meistrich ML, Trostle PK, Brock WA (198t) Association of nuc-

leoprotein transitions with chromatin changes during rat spermatogenesis. In: Jagiello G, Vogel HJ (eds) 'Bioregulators of reproduction'. Academic New York, pp 151-166

Miller OL, Bakken AH (1972) Morphological studies on transcription. Acta Endocrinol 168:155-173

Oudet P, Gros-Bellard M, Chambon P (1975) Electron microscopic and biochemical evidence that chromatin structure is a repeating unit. Cell 4:281-300

Platz RD, Grimes SR, Meitrich ML, Hnilica LS (1975) Changes in nuclear proteins of rat testis cells separated by velocity sedi- mentation. J Biol Chem 250:5791-5800

Pogany GC, Corzett M, Weston S, Balhorn R (1981) DNA and protein content of mouse sperm. Implications regarding sperm chromatin structure. Exp Cell Res 136:127-136

Sautiere P, Belaiche D, Martinage A, Loir M (1984) Primary structure of the ram (Ovis aries) protamine. Eur J Biochem 144:121-125

Scheer U (1978) Changes of nucleosome frequency in nucleolar and non-nucleolar chromatin as a function of transcription: An electron microscopic study. Cell 13:535-549

Suau P, Subirana JA (1977) X-ray diffraction studies of nucleoprotamine structure. J Mol Biol 117 : 909-926

Tsanev R, Avramova Z (1981) Non-protamine nucleoprotein ultra- structures in mature ram sperm nuclei. Eur J Cell Biol 24:139-145

Wagner TE, Yun JS (1981) Human sperm chromatin has a nucleosomal structure. Arch Androl 7 : 251-257

Wagner TE, Sliwinski JE, Schewmaker DB (1978) Subunit structure of eutherian sperm chromatin. Arch Androl 1:31-41 Young RJ (1979) Rabbit sperm chromatin is decondensed by a

thiol-induced proteolytic activity not endogenous to its nucleus. Biol Reprod 20:1001-1004

Young R J, Sweeney K (1979) The structural organization of sperm chromatin. Gamete Res 2:265-283

Zirkin BR, Chang TSK, Heaps J (1980) Involvement of an acrosin- like proteinase in the sulfhydryl-induced degradation of rabbit sperm nuclear protamine. J Cell Biol 85:116-121