Biochimica et Biophysica Acta, 434 (1976) 377-389

© Elsevier Scientific Publishing Company, Amsterdam - - Printed in The Netherlands BBA 37355 COMPARATIVE ANALYSIS OF T H E N U C L E A R BASIC PROTEINS IN RAT, H U M A N , G U I N E A PIG, MOUSE A N D RABBIT SPERMATOZOA

HAROLD I. CALVIN International Institute for the Study of Human Reproduction, Columbia University College of Physician~ and Surgeons, New York, N.Y. 10032 (U.S.A.)

(Received December 31st, 1975)

SUMMARY Cysteine-rich protamines (Arg = 47-61 Yoo;Cys = 8-16 7o) were isolated from the sperm of an individual guinea pig, human and rabbit and from pooled samples of mouse and rat sperm. Appreciable concentrations of histones were not found in the sperm nuclei of these species. In addition to the protamines, a substance of relatively low molecular weight, which reacted with the Lowry reagent, appeared in crude acid; soluble extracts of sperm nucleoprotein. This unidentified contaminent was resolved from the protamines by chromatography on Bio-Rex 70. Heterogeneity of human and mouse protamines was revealed by electrophoresis at pH 2.7, in the presence of 2.5 M urea, and confirmed by amino acid analysis, which also suggested the presence of 2 or more species of protamine in the rabbit. By contrast, the guinea pig and rat preparations displayed nearly stoichiometric ratios of amino acid residues, approaching homogeneity by this criterion. The functional consequences of crosslinks between cysteine residues of these proteins and the possible species-specific significance of their differing percentages of histidine are discussed. Potentially analogous functions are suggested for phosphorylated serine and threonine, and for ionized cysteine and tyrosine, within the protamines of developing spermatids. Their amino acid compositions indicate that the protamines of eutherian mammals are coded by a C. G-rich genome which has been unusually susceptible to genetic drift. An especially high rate of G -~ A transitions seems to have occurred in the human protamine genes.

INTRODUCTION The presence of a relatively high content of protein-bound cystine in mammalian spermatozoa was first recognized over 30 years ago [1, 2], suggesting that the unusual structural stability of these cells could be ascribed to -S-S- bridges between 1/2-cystine residues of neighboring protein molecules. Localization of cystine-rich proteins within the chromatin of the sperm nucleus was later indicated by studies of the conditions required for extraction of DNA [3] and protein [4, 5] from the nucleus

378 and by ultrastructural observations of nuclear swelling in the presence of disulfidereducing agents [6]. Eventually, the isolation and preliminary characterization of a single cysteinerich basic protein from bull sperm heads was accomplished by Coelingh et al. [7], following reduction of -S-S- bonds with mercaptoethanol, in the presence of guanidinium chloride, and aminoethylation of the liberated -SH groups. Subsequently, Rozijn's group has reported the complete amino acid sequence of this molecule [8] and the compositions and partial sequences of similar proteins isolated from ram, boar and stallion sperm [9]. In addition, amino acid analyses of several other mammalian protamines have been obtained in our laboratory [10] and elsewhere [11-15]. In all cases, the isolated basic proteins were found to be rich in arginine, but contained relatively little lysine and were obviously of a lower molecular weight than histones. They have thus been classified as protamines [10, 16, 17], no doubt phylogenetically related to those of teleost and avian sperm. Their high content of cysteine, however, distinguishes them from the protamines of lower vertebrates and has encouraged their designation as a special subclass, the cysteine-protamines [16] or thiomines [10]. The interest of this investigator in such proteins stems from morphological studies which indicated that a high concentration of-S-S- crosslinks within the sperm nucleus, established mainly during passage of the sperm through the epididymis, is probably universal among eutherian mammals [18], although such crosslinks cannot be detected in marsupials [18], and are either totally absent or present to a more limited degree in the sperm of a variety of submammalian species [19]. As a corollary of these studies, the basic nuclear proteins of guinea pig, man, mouse, rabbit and rat sperm have been isolated and partially characterized. Analyses of human, rabbit and rat protamines have already been reported without a detailed account of methods [10]. In the present communication, the procedures for isolation of these and two other sperm basic nuclear protein fractions, those of the guinea pig and mouse, are described and the preparations characterized by electrophoresis and amino acid analysis. The amino acid compositions of various eutherian protamines are discussed from the point of view of molecular evolution and function. MATERIALS AND METHODS

Spermatozoa. Rat (Sprague-Dawley) and mouse (Swiss-Webster, random-bred) testes, with epididymides attached, purchased from Pel-Freez (Little Rock, U.S.A.), were shipped on ice, and maintained at 0-2 °C until use. The spermatozoa were expressed from the cauda of the epididymis within 3-4 days of sacrifice and filtered to remove tissue debris. Cauda epididymal spermatozoa were obtained similarly, immediately after sacrifice, from a 12-1b New Zealand white rabbit (Wallingford Laboratories, Wallingford, U.S.A.). Human spermatozoa, derived from the semen of a healthy donor, was washed several times by dilution with 0.02 M NaH2PO4, pH 6.0, and centrifugation at 1500 × g for 10 min. Preparation of chromatin. To dissolve both the tails and the perinuclear structures within the sperm heads, guinea pig, human, mouse and rabbit spermatozoa were incubated for 20 min at 25 °C in 8 M urea/2 mM dithiothreitol/50 mM Tris. HC1, pH 9.0, and sonicated for 30 s at 55 W with a Bronwill Biosonik Model IIA sonifier,

379 fitted with a 5/32-inch microtip. Aliquots were removed for the counting of sperm nuclei with a hemocytometer [20], the sample was cooled to 5 °C, and the chromatin collected by centrifugation at 10 000 x g for 10 min. Rat chromatin was prepared similarly, except that a higher concentration of dithiothreitol (5 mM) was required to dissolve the tails. In addition, both rat and human chromatin were prepared in urea/ dithiothreitol from isolated sperm heads [20], as well as from intact spermatozoa. All chromatin pellets were dissolved immediately, as described below, to avoid prolonged exposure to urea/dithiothreitol. Solubilization of chromatin and aminoethylation. The procedures for solubilization and aminoethylation were modified slightly from those of Coelingh et al. [7]. Chromatin pellets were suspended in water (0.5-1.0"109 sperm nuclei/ml) and sonicated at 55 W until evenly suspended. The material was delivered dropwise, with a disposable Pasteur pipet whose tip had been drawn out (bore diameter: 0.1-0.2 mm), into 20 vol. of a rapidly stirring solution of 5.0 M guanidinium chloride/10 mM dithiothreitol/0.20 M Tris. HC1, pH 9.0, over a period of 5-10 min. Following an additional 30 min of incubation at 25 °C, with stirring, the clear viscous solution was diluted with 0.30 vol. 3.0 M Tris.HCl, pH 8.6, and submitted to aminoethylation with ethylenimine, as described by Cole [21]. The disappearance of-SH was monitored with Ellman's reagent [22]. When -SH was no longer detected (30-40 min), the solution was cooled and adjusted below pH 3.0 with concentrated HCI. Isolation of nuclear basic proteins. Dialysis was performed in cellulose tubing (1 inch flat width, 0.0008 inch thick), purchased from A. H. Thomas (Philadelphia, U.S.A.), which was acetylated (without prestretching) at 65 °C to reduce pore size [23]. The aminoethylated chromatin was dialyzed for 3 days against 3 changes of 50-100 vol. 0.25 M HCI and the precipitated DNA eliminated by centrifugation. Following 6-8 h of further dialysis against water, the extract was concentrated 10-20-fold by flash evaporation, analyzed for protamine by the Sakaguchi reaction [24], neutralized to pH 3.0 by the addition of conc. NazCO3, mixed with acetic acid and NaC1 to a final composition of 0.2 N acetic acid/1.0 M NaC1, and applied to a column of BioRex 70 (BioRad, Richmond, U.S.A.), equilibrated to pH 3.0 with the same medium. Elution was continued with 0.2 M acetic acid/1.0 M NaCI (1.5 column vol.), 0.2 M acetic acid (1.5 column vol.) and 0.25 M HC1 (2.0 column vol.). The fractions were assayed for ultraviolet absorption at 260 and 280 nm and the ultraviolet-positive fractions assayed colorimetrically by the Lowry et al. [25] and Sakaguchi [24] reactions. Assay of protein. The reaction of Lowry et al. [25] was found to be applicable only for the determination of purified proteins, since a low molecular weight, positively reacting material of unknown composition appeared in crude acidic extracts of sperm chromatin. The concentration of protamine in both crude extracts and purified fractions was estimated with a modified Sakaguchi reagent [24]. Sample dilutions were prepared in 0.5 M NaOH, instead of in water. This additional modification provided increased sensitivity and facilitated the assay of samples dissolved in dilute HC1. The assay was calibrated with salmon protamine (Schwarz-Mann, Orangeburg, U.S.A.). Inhibition of color development by contaminants such as guanidinium chloride was readily detected by measuring absorption at 450 nm, as well as at 525 nm. A relatively low absorption at 450 nm was characteristic of inhibited samples, often negative with respect to the blank.

380

Electrophoresis. Samples of crude or purified basic protein were electrophoresed in cylindrical 5 mm x 60 mm 15~ acrylamide gels for 2.5 h at 2.5 mA per gel, in the presence of 0.37 M glycine, pH 4.0, in parallel with standards of salmon protamine (Schwarz-Mann, Orangeburg, U.S.A.) and calf thymus histone (Worthington Biochemical, Freehold, U.S.A.), and stained with Amido Black, following the procedure of McAllister et al. [26]. Alternatively, the samples were electrophoresed in 15 ~ gels containing 2.5 M urea, in the presence of 0.9 M acetic acid, as described by Panyim and Chalkley [27]. Following pre-electrophoresis at 1.8 mA per gel for 4 h, the samples were applied to the gels, electrophoresed for 15 min at 0.9 mA per gel and then for 75 min at 1.8 mA per gel. For detection of the proteins, the gels were pretreated for 30 rain with 0.1 M sodium picrate, pH 7.0, rinsed briefly with water and then stained overnight with 0.15~o Coomassie Brilliant Blue (Schwarz-Mann, Orangeburg, U.S.A.), dissolved in 50 ~ methanol/7 ~ acetic acid. Destaining was carried out by passive diffusion in 5 ~ methanol/7 ~ acetic acid. The quantities of protamine or of total acid-soluble extract applied to the gels were estimated by the Sakaguchi reaction [24]; all other samples were estimated by weight or by the Lowry et al. reaction [25]. Amino acid composition. Samples to be analyzed for amino acid composition were lyophilized in ampoules, dissolved in 6 M HC1, and equilibrated with Nz. The ampoules were sealed under a vacuum of less than 25 #m and incubated for 24 h at ll0-112°C. The hydrolyzate was lyophilized and analyzed with a Model 120 analyzer (Beckman Instruments, Palo Alto, U.S.A.) on 1 × 50 cm columns of AA-15 resin (Beckman Instruments, Palo Alto, U.S.A.), using the 2-column system of Spackman et al. [28]. Approx. 500-800 #g of protein was employed for a single analysis. No corrections were made for losses during hydrolysis. RESULTS

Electrophoresis of sperm nuclear basic proteins The crude acid-soluble extracts of the dissolved nucleoproteins, prior to chromatography on Bio-Rex 70, displayed two bands upon electrophoresis at pH 4.0. The presence of these two components in human sperm nucleoprotein is illustrated in Fig. la. A similar electrophoretic pattern was obtained with corresponding extracts prepared from rat, mouse, rabbit and guinea pig sperm. In each case, the more rapidly migrating component, evidently of relatively low molecular weight and highly basic, was not adsorbed by Bio-Rex 70 in the presence of 0.2 M acetic acid/1.0 M NaC1. The fractions in which it appeared did not display the presence of arginine, but did yield a strongly positive assay with the Lowry et al. reagent. The chemical nature of this low molecular weight contaminant has not been established. Its presence in the sperm head must remain in question until the possibility of artifact is eliminated by varying the procedures for preparation of sperm nucleoprotein. During chromatography on Bio-Rex 70, the cysteine-rich protamines of the sperm head were quantitatively adsorbed in 0.2 M acetic acid/1.0 M NaC1 and could then be eluted free of the above contaminant by rinsing the column with dilute HCI. This chromatographic purification was verified by electrophoretic analysis of the two peaks of ultraviolet-absorbing material eluted from the column. Fig. lb and c illustrates the electrophoretic behavior of peaks I and II obtained from a human

381

Fig. 1. Polyacrylamidegel electrophoresis in 15 ~ gels at pH 4.0 [26] of fractions derived from human sperm nucleoprotein. An internal standard of methyl green (5 #g) was coelectrophoresed with each sample. Stained with Amido Black. (a) Total acid-solubleextract (8/~g) [24], (b) peak I from Bio-Rex 70 chromatogram (20/~g) [25], (c) peak II from Bio-Rex 70 chromatogram (10/tg) [24], (d) salmon protamine (10¢tg). The faint line in gels c and d, which appears to coincide with peak I, is probably derived from methyl green. sperm acid-soluble extract. Peak I coincided with the more rapidly migrating component, whose mobility was considerably greater than that of salmon protamine (Fig. ld), whereas Peak lI, the human protamine fraction, migrated somewhat more slowly than salmine. Following purification by chromatography on Bio-Rex 70, the basic nuclear proteins of rat, human, mouse, rabbit, and guinea pig sperm displayed a single band upon electrophoresis in 15 ~o polyacrylamide gels at pH 4.0 [26], although occasionally two components of mouse basic protein were revealed by this technique. These results were consistent with the assumption that either a single basic protein or else a narrow class of basic proteins is present in the sperm nuclei of eutherian mammals. To investigate further the possible heterogeneity of these proteins, the samples were electrophoresed at pH 2.7 in the presence of 2.5 M urea [27], using a novel staining technique in which the gels were pretreated with sodium picrate before staining with Coomassie Blue. This highly sensitive detection procedure enabled the demonstration of varying degrees of heterogeneity of the basic protein fractions, most notably in mouse sperm (Fig. 2a). tn the mouse, two major protamine bands, as well as trace components of greater electrophoretic mobility, were detected. Mouse protamine has been found by others to display two bands [29] or one broad band [30] when electrophoresed under similar conditions. A slight degradation of mouse protamine, during incubation of sperm heads in 8 M urea/2 mM dithiothreitol at pH 9.0, is postulated on the basis of Fig. 2a and a recent report that prolonged incubation of bull sperm chromatin, in the presence of 0-4 M urea and 0.02-0.2 M mercaptoethanol at pH 8.0, results in extensive

382

Fig. 2. Electrophoresis of mammalian protamines in 15 ~ gels containing 0.9 M acetic acid and 2.5 M urea [27]. Fixed in picrate and stained with Coomassie Blue. Gels a-f contain 6-10/~g of protamine [24], purified by chromatography on Bio-Rex 70 or prepared as crude acid-soluble fraction of nucleoprotein derived from sperm heads pretreated with urea/dithiothreitol: (a) mouse, purified; (b) human, crude; (c) human, purified; (d) rabbit, purified; (e) rat, purified; (f) guinea pig, purified. Gel g contains 10/~g of calf thymus histone. proteolysis of the nuclear basic protein [31]. The latter observations were published after the work reported here had already been completed. Fortunately, proteolysis does not seem to have occurred in the other samples analyzed in the present study (Fig. 2b-f). Nevertheless, preliminary observations have indicated that the sperm nuclei of these species, when treated with urea/dithiothreitol, become grossly distorted and begin to decondense as a result of prolonged incubation, storage or freezethawing. Moreover, a slight proteolysis of bull sperm protamine has been detected in this laboratory by polyacrylamide gel electrophoresis, following 20 min incubation in 8 M urea/2 mM dithiothreitol at pH 9.0. It is therefore recognized that treatment of spermatozoa with urea/dithiothreitol, although a convenient method to selectively dissolve the tails, is not a sound procedure for purification of chromatin. Isolation of sperm heads, followed by detergent treatment [20, 31], is preferable. The apparent multiplicity of human protamines, which has been reported by other laboratories [12, 15], is especially evident in the gel shown in Fig. 2b, in which an acid-soluble extract of chromatin purified in urea/dithiothreitol was electrophoresed. Three bands are detectable in the protamine zone. More typically, under the conditions of this experiment, partial resolution of the major band does not occur, due to less ideal zone sharpening at the origin, and only two bands are evident, as illustrated by the pattern shown in Fig. 2c, obtained with chromatographically purified human protamine. In contrast to those of human and mouse, the sperm nuclear basic protein fractions of rabbit (Fig. 2d), rat (Fig. 2e) and guinea pig (Fig. 2f) did not display more than one discrete band in the protamine region of the gel, although the single

383 protamine component appeared to be diffuse, especially in the case of the guinea pig. No attempt was made to eliminate phosphate groups from these protamines. These have been shown to occur in rat protamine isolated from mature sperm, which forms a broad band during electrophoresis unless treated with phosphatase [32]. To investigate the reported possibility of histones in the human sperm head [12, 15], nucleoprotein extracts, prepared in the usual way from chromatin purified in urea/dithiothreitol (Fig. 2b) or from sperm heads isolated on sucrose gradients following sonication (Fig. 3a and b), were electrophoresed at pH 2.7 in the presence of 2.5 M urea and the gels treated successively with picrate and Coomassie Blue. In spite of the great sensitivity of this technique, bands with the mobility of histones (Figs. 2g and 3c) were either barely detectable in the presence of moderate quantities of protamine (Figs. 2b and 3a), or else evidently of marginal significance in those gels which were overloaded with basic protein (Fig. 3b). Although traces of histone-like material have been found in other eutherian protamine fractions (e.g. rabbit, Fig. 2d), these do not occur consistently and thus may be derived from contaminating somatic cell nuclear chromatin.

Fig. 3. Electrophoresis in 0.9 M acetic acid/2.5 M urea of acid-soluble extract of nucleoprotein prepared from human sperm heads: (a) 5/~g of crude protamine, (b) 30/~g of crude protamine. Gel (c) contains 10/~g of calf thymus histone. In this run, the proteins migrated somewhat further than in the experiments depicted in Fig. 2. Amino acid compositions The amino acid compositions determined in this laboratory on samples of guinea pig, human, mouse, rabbit and rat sperm head basic proteins are listed in Table I. The analyses of the human and rabbit samples, each derived from the sperm of a single male, suggest heterogeneity of the basic protein in the sperm head. Assuming a polypeptide chain of approx. 50 residues, which appears to be characteristic of mammalian protamines, the presence of amino acid residues in concentrations

384 TABLE I AMINO ACID COMPOSITIONS OF MAMMALIAN SPERM NUCLEAR BASIC PROTEINS The data are expressed in mol ~ and are based on 2 4 determinations, except for the guinea pig data, which is based on a single analysis. Concentrations of less than 0.4 mol ~ have been omitted. Amino acid

Guinea pig

Human ~

Mouse

Rabbit

Rat*

Ala Arg Asp Cys Glu Gly His

1.8 (1)** 59.6 (30) -12.0 (6) -1.2 3.8 (2) 0.4 0.4 1.9 (1) -1.9 (1) 1.8 (1) 6.5 (3) 1.9 (1) 7.6 (4) 0.4

2.0 46.8 8.5 6.7 2.6 9.4

1.2 52.6 0.5 10.5 1.0 3.5 12.7

54.4 -13.9 4.6 0.8 0.7

1.9 (1)** 61.2 (30-31) -15.3 (8) -0.5

1.0

1.7

1.1 2.5 0.4 . . 1.9 8.3 3.4 4.4 --

1.3 5.2 -. . .

lie Leu

Lys Met Phe Pro Set Thr Tyr Val

.

2.2 2.4 --

8.1 1.1 2.2 --

--

--

6.0 4.1 5.6 4.5

-4.1 (2) -1.9 (I) -7.7 (4) 1.9 (l) 5.8 (3) --

* Duplicate determinations were carried out on each of two samples, derived respectively from chromatin prepared by two slightly different procedures (see text). ** Values in parentheses refer to the number of residues/molecule represented by the data, assuming that 2 mol ~ = 1 residue/molecule (see text). which do n o t a p p r o x i m a t e 2 mol 9/00, or a m u l t i p l e thereof, is indicative o f this fact. By this measure, the m o u s e p r o t a m i n e , isolated f r o m a large n u m b e r o f animals, is also clearly heterogeneous, in a g r e e m e n t with its e l e c t r o p h o r e t i c behavior. On the o t h e r hand, r a t p r o t a m i n e , i s o l a t e d f r o m a p o o l o f males, and guinea pig p r o t a m i n e , isolated f r o m a single male, a p p e a r to a p p r o a c h homogeneity. The basic p r o t e i n fractions whose analyses are listed in Table I contain 47-61 ~ arginine, 8-16 9/o cysteine and 6-8 9/00serine. Relatively little lysine is present (2-5 9/00). O t h e r n o t e w o r t h y features are the absence o f a s p a r t a t e (or asparagine) and the widely-varying content o f tyrosine, histidine and g l u t a m a t e (or glutamine) in these proteins. The analysis o f m o u s e s p e r m nuclear basic p r o t e i n indicates that, c o n t r a r y to a p r e v i o u s r e p o r t [33], this fraction is relatively low in lysine a n d is at least m a i n l y a mixture o f cysteine-rich p r o t a m i n e s . DISCUSSION The e l e c t r o p h o r e t i c b e h a v i o r o f the five m a m m a l i a n s p e r m basic protein p r e p a r a t i o n s which were a n a l y z e d in the present study suggests t h a t the m a j o r species in each p r e p a r a t i o n are o f similar m o l e c u l a r weight, b u t t h a t the degree o f h o m o geneity varies, the m o u s e a n d h u m a n fractions being m o r e c o m p l e x than those o f the rat, r a b b i t or guinea pig. The e l e c t r o p h o r e t i c heterogeneity o f mouse [29] and h u m a n [12, 15] p r o t a m i n e s in the system o f P a n y i m a n d C h a l k l e y [27] has been r e p o r t e d b y others, whereas the p r o t a m i n e s in rat [32, 34], r a b b i t [29, 34] or guinea

385 pig [29] sperm have been found to display a single band. In the rat, it has been demonstrated that this band is broadened somewhat by the minor degree of phosphorylation which remains in the mature protamine [32]. In addition, Kumaroo et al. [13] have reported the presence of a second, slightly more rapidly migrating component in rat sperm basic protein. The amino acid composition of this minor component includes a high percentage of arginine and essentially no lysine or cysteine. No evidence for the existence of such a protein has been noted in the present investigation or in several other laboratories [11, 14, 32, 34]. Further studies would be necessary to confirm its association with the sperm nucleus and determine its quantitative significance. In any case, the nuclear basic protein of rat sperm is comprised almost entirely of a cysteine-rich protamine, whose amino acid composition deserves some comments. The molar percentage of cysteine listed in Table I for this protein is an average of four determinations. It was, in addition, verified on carboxamidomethylated protein and found to be the same, after correcting for loss of carboxymethylcysteine during hydrolysis ( < 10~ for a standard of carboxymethylcysteine incubated under the same conditions). Others have reported considerably lower [11, 13], or slightly lower [14] percentages of this amino acid. It is probable that these lower estimates reflect incomplete reduction of -S-S- bonds or destruction of labile -SH derivatives (particularly carboxymethylcysteine) during hydrolysis. It is especially critical to dissolve sperm nucleoprotein completely when reducing -S-S- bonds. The approach used in our laboratory involves dropwise addition of small quantities of sperm chromatin into the dissolving medium to minimize the formation of stubborn aggregates of swollen nucleoprotein. Analysis of the mouse nuclear basic protein fraction suggests considerable heterogeneity. In addition, the 0.5 m o l ~ content of aspartate raises the possibility of slight contamination by material not derived from the sperm nucleus. Nevertheless, the fraction is essentially protamine, as evidenced by its high percentage of arginine (53 ~) and modest percentage of lysine (5 ~). The lysine and arginine-rich protein reported by Lam and Bruce [33], if present, is a minor constituent of this fraction. Kistler et al. [ll] have found a similar protein in rat testis which does not appear in the sperm and is thought to be associated with the developing spermatid nucleus. The variable stability of human sperm heads within a single ejaculate, a relatively rare phenomenon not observed in other eutheria, may be explained at least partially by the isolation of two arginine-rich proteins of differing cysteine content and quite dissimilar amino acid composition from pooled human ejaculates by Kolk and Samuel [12]. Since a product of intermediate composition was isolated from the sperm of a single individual in the present study, both varieties of proteins characterized by these workers may well be present in every human sperm head. The variable swelling resistance of the sperm nuclei within a single ejaculate could depend on the proportion of these two types of protamine in each sperm head. In addition to their moderate difference in cysteine content, other dissimilarities in composition, most notably their relative percentages of histidine, could of course affect the comparative packing behavior of the two protamines within the human sperm nucleus. The reports of significant concentrations of histone-like proteins in the human sperm head [12, 15] are in conflict with this investigator's electrophoretic analysis of acid-soluble extracts derived from chromatin purified in urea/dithiothreitol (Fig. 2b)

386 or from whole sperm heads (Fig. 3a and b). Since the histone-like material reported by other laboratories was designated as such largely on the basis of crude chromatographic separations and electrophoresis at pH 3.0, it is questionable that these proteins are truly histones or that they are even associated with the nucleus. In one of the procedures [12], the nucleoprotein was prepared from whole sperm without removal of acid-insoluble proteins. In the other study [15], precipitation ot DNA and non-histone proteins was conducted without removal of guanidinium chloride by dialysis, permitting the possible solubilization of proteins which would ordinarily precipitate with DNA at pH < 1.0. Analysis of the "histone" fraction isolated from the latter extracts did reveal a considerable similarity to liver histone, but showed a significantly higher ratio of [Asx + Glx] : [Lys + His + Arg] (1.45 vs. 0.96). On the basis of the amino acid analyses presented here and elsewhere, it is apparent that the evolution of mammalian protamines has been exceptionally nonconservative. This characteristic, which is true of the vertebrate protamines in general, suggests a relatively non-specific function for these molecules [35]. Nevertheless, certain amino acid residues are consistently prominent in the protamines of eutherian mammals, according to the data shown in Table I and results obtained by others [9,12]: (1) arginine (44-61 ~), (2) cysteine (8-16~) and (3) serine plus threonine (8-13 ~). The relative proportions of these residues may have an important bearing on the function of protamines during differentiation of the sperm nucleus. Extensive ionization of cysteine, favored by the proximity of the highly basic arginine residues, as well as phosphorylation of serine (and probably threonine)[32, 36], both occur within the protamines of maturing spermatids. These negative charges are lost during sperm maturation, respectively by oxidation of cysteine to cystine [18, 32], and by dephosphorylation of serine (or threonine) [32], thereby restoring the full basicity of the protamine which is essential to its stable association with DNA. It has been suggested that the phosphate groups in protamines of developing spermatids are required during transport of the basic proteins from their site of synthesis into the nucleus [36] or during the process of nuclear condensation [32]. Similar functions might exist for other ionizable groups, such as the -SH in cysteine and perhaps also the -OH in tyrosine. The high content of disulfide crosslinks engendered by the oxidation of cysteine residues in the basic proteins of eutherian mammals is exceptional and of profound consequence for the structural and chemical stability of the nucleoprotein. It has been proposed by us that such stabilization may serve to exclude potential mutagens from the nucleus [37] or, more likely, to assist in penetration of the zona pellucida during fertilization [19]. More recently, evidence has been presented that -S-S- bonds protect the chromatin from autolysis [31]. In any case, the contents of cysteine in the protamines, which range between 8 and 16 ~o, distinguish them from those of teleost and avian sperm, which do not contain cysteine (see Bloch [35] and Coelingh and Rozijn [17]). The introduction of cysteine residues, and consequently disulfide bands, into the sperm nucleus must have required the prior existence of the mechanisms to prevent the premature formation of-S-S- crosslinks between protamine molecules before completion of their assembly into the chromatin, as well as agents to bring about the disruption of these bonds during decondensation of the sperm nucleus in the fertilized egg. The most likely candidates for the latter function are thiol-rich substances, known to be present in high concentration in mammalian oocytes [38].

387 It is possible to draw certain conclusions concerning the molecular evolution of the protamines of eutherian mammals by considering which amino acids other than arginine appear in relatively high concentrations in one or more eutherian protamines. To simplify the argument, it will be assumed that all of the glutamate in the protamine hydrolyzates is derived from glutamine residues. This is not an unlikely possibility, since neither glutamate nor aspartate has ever been demonstrated to occur in vertebrate protamines, whereas glutamine has been identified in thynnine Y2, and in ram and bull protamines (see Coelingh and Rozijn [17]). Moreover, those organically bound, negatively charged groups which occur in the protamines of immature sperm, i.e. phosphate and sulfide, are almost entirely removed or transformed during sperm maturation, as discussed previously, suggesting that negative charges are incompatible with the function of protamine in fully mature chromatin. As many as five different amino acids occur at least four times in one or more of the eutherian protamine molecules, on the basis of a survey of data from this (Table I) and 2 other laboratories [9, 12]. These are cysteine, histidine, serine, tyrosine and glutamine (assuming that glutamate is forbidden in protamines). The implications of this information are summarized in Table II. There are six codons for arginine: CGA, C G G , CGC, C G U , A G A and A G G . Either cysteine, histidine, serine or glutamine can replace arginine in protamines by means of a single point mutation of one of the four codons whose first two nucleosides are C and G, i.e. CGX. Only serine can enter the molecule by a single point mutation from one of the A G codons and only tyrosine cannot arise by any single-step mutation from arginine. TABLE II POTENTIAL REPLACEMENTS OF ARGININE BY VARIOUS FREQUENTLY OCCURRING RESIDUES IN EUTHERIAN PROTAMINES Amino acid

Max. no. residues per molecule

Codons

CGX*

AG (A, G)**

C --~ U G~ A None

None None None

AG(U,C) U A (U, C)

C-+A None

(A, G) ~ (U, C) None

CA (A, G) GA (A, G)

G --+ A None

None None

Cys His Ser

8 (boar) [9], (rat)*** UG (U, C) 8 (human 2) [12] CA (U, C) 5 (human 1) [12] UCX

Tyr

4 (guinea pig)***

Gln or Glu

(human 1) [12] 4 (human 1) [12]

Single-stepmutation from

* The letter X signifies the presence of four possible codons, ending respectively in A, G, C or U. ** This nomenclature is used to signify the presence of two alternate codons. *** Derived from analysis listed in Table I. Other values obtained from the references indicated. The above analysis supports the contention of Ling and Dixon [39] and of Coelingh and Rozijn [17] that the sequence C G predominates in the R N A which codes for the synthesis of vertebrate protamines. It seems likely, moreover, that the residues listed in Table II, with the notable exception of tyrosine, have usually entered the protamines by mutation of one of the C G X codons for arginine.

388 The highly variable content of histidine in eutherian protamines is especially interesting, for this amino acid, though present at a concentration of only 0-1 residue per molecule in human protamine 1 [12], and in the protamines of four ungulates [9], the rabbit (Table I) and the rat (Table I), is a major constituent of mouse protamine (approx. 6 residues/molecule) (Table I) and human protamine 2 (8 residues/molecule) [12]. The large difference in histidine content between protamine 1 (1 residue/ molecule) and protamine 2 (8 residues/molecule) in the human is especially striking and is probably explained by gene duplication, followed by a series of G ~ A point mutations in the CGU and CGC codons for arginine. Similarly, G -+ A transitions could have replaced four arginines by glutamines in human protamine 1 (Table II). The apparently high rate of such mutations in the human protamine correlates well with their relatively high frequency in human hemoglobin genes [40], which has suggested the possibility of a specific mechanism in humans for the catalysis of G -+ A. Contrasting frequencies of histidine in various eutherian protamines suggest the possibility that its presence may be related to a species-specific feature of the sperm nucleus, such as its shape. It is especially noteworthy that protamines derived from flattened, spatulate sperm nuclei display a low content of histidine (0-1 residues/ molecule in bull, boar, ram, stallion and rabbit). On the other hand, mouse and human sperm nuclei, whose shapes are more distinctive and clearly different from those of the other species so far analyzed, contain at least one protein relatively rich in histidine. Thus, in this respect at least, the hypothesis that protein composition significantly influences the form of the sperm nucleus [41] may be borne out. Whatever the roles of protamines in the specific morphogenesis of eutherian sperm nuclei, these apparently do not place highly rigid constraints on amino acid composition, for there is considerable variation in composition among ungulate protamines [9]. No doubt the correlation between structure and function in the protamines is relatively flexible, the requirements being considerably less specific than in enzymes, binding proteins or histones. The observation that non-arginine residues are clustered primarily at the ends of the bull protamine [8] suggests the possibility of a functional dichotomy between the ends and middle of the eutherian protamine molecule. Sequence studies involving additional species might further clarify the constraints which function places upon the primary structure of protamines in eutherian mammals. ACKNOWLEDGMENTS The author wishes to thank Miss Herma Wohlrab and Mrs. C. C. Yu for technical assistance, Mr. Walter Schrepel for performing the amino acid analyses and Dr. Allen Gold for valuable discussions. This work was supported by grant no. HD05316 from the National Institutes of Health. REFERENCES 1 Green, W. W. (1940) Anat. Rec. 76, 455-471 2 Zittle, C. A. and O'Dell, R. A. (1941) J. Biol. Chem. 140, 899-907 3 Borenfreund, E., Fitt, E. and Bendich, A. (1961) Nature 191, 1375-1377 4 Bril-Petersen, E. and Westenbrink, H. G. K. (1963) Biochim. Biophys. Acta 76, 152-154 5 Henricks, D. M. and Mayer, D. T. (1965) Exp. Cell Res. 40, 402-412

389 6 Lung, B. (1968) J. Ultrastruct. Res. 22, 485-493 7 Coelingh, J. P., Rozijn, T. H. and Monfoort, C. H. (1969) Biochim. Biophys. Acta 188, 353-356 8 Coelingh, J. P., Monfoort, C. H., Rozijn, T. H., Leuven, J. A. G., Schiphof, R., Steyn-Parve, E. P., Braunitzer, G., Schrank, B. and Ruhfus, A. (1972) Biochim. Biophys. Acta 285, 1-14 9 Monfoort, C. H., Schiphof, R., Rozijn, T. H. and Steyn-Parve, E. P. (1973) Biochim. Biophys. Acta 322, 173-177 10 Calvin, H. I. (1975) in Biology of the Male Gamete (Duckett, J. G. and Racey, P. A., eds.), pp. 257-273, Academic Press, London 11 Kistler, W. S., Geroch, M. E. and Williams-Ashman, H. G. (1973) J. Biol. Chem. 248, 4532-4543 12 Kolk, .~. M. and Samuel, T. (1975) Biochim. Biophys. Acta 393, 307-319 13 Kumaroo, K. K., Jahnke, G. and Irvin, J. L. (1975) Arch. Biochem. Biophys. 168, 413-424 14 Platz, R. D., Grimes, S. R., Meistrich, M. L. and Hnilica, L. S. (1975) J. Biol. Chem. 240, 57915800 15 Puwaravutipanich, T. and Panyim, S. (1975) Exp. Cell Res. 90, 153-158 16 Subirana, J. A., Cozcolluela, C., Palau, J. and Unzeta, M. (1973) Biochim. Biophys. Acta 317, 364-379 17 Coelingh, J. P. and Rozijn, T. H. (1975) in The Biology of the Male Gamete (Duckett, J. G. and Racey, P. A., eds.), pp. 245-256, Academic Press, London 18 Calvin, H. I. and Bedford, J. M. (1971) J. Reprod. Fertil. Suppl. 13, 65-75 19 Bedford, J. M. and Calvin, H. I. (1974) J. Exp. Zool. 188, 137-155 20 Calvin, H. I. (1976) Methods Cell Biol. 13, 85-104 21 Cole, R. D. (1967) Methods Enzymol. 11, 315-317 22 Ellman, G. L. (1959) Arch. Biochem. Biophys. 82, 70-77 23 Craig, L. C. and Konigsberg, W. (1961) J. Phys. Chem. 65, 166-172 24 Dubnoff, J. W. (1957) Methods Enzymol. 3, 635-639 25 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275 26 McAllister, H. C., Wan, Y. C. and Irvin, J. L. (1963) Anal. Biochem. 5, 321-329 27 Panyim, S. and Chalkley, R. (1969) Arch. Biochem. Biophys. 130, 337-346 28 Spackman, D. H., Stein, W. H. and Moore, S. (1958) Anal. Chem. 30, 1190-1206 29 Bellve, A. R. and Romrell, L. J. (1974) J. Cell Biol. 63, 19a (abstr) 30 Ecklund, P. S. and Levine, L. (1975) J. Cell Biol. 66, 251-262 31 Marushige, Y. and Marushige, K. (1975) Biochim. Biophys. Acta 403, 180-191 32 Marushige, Y. and Marushige, K. (1975) J. Biol. Chem. 250, 39-45 33 Lam, D. M. K. and Bruce, W. R. (1971) J. Cell. Physiol. 78, 13-24 34 Shires, A., Carpenter, M. P. and Chalkley, R. (1975) Proc. Natl. Acad. Sci. U.S. 72, 2714-2718 35 Bloch, D. P. (1969) Genetics, Suppl. 1, 93-111 36 Ingles, C. J. and Dixon, G. H. (1967) Proc. Natl. Acad. Sci. U.S. 58, 1011-1018 37 Calvin, H. I. and Bedford, J. M. (1974) J. Reprod. Fertil. 36, 225-229 38 Barrnett, R. J. (1953) J. Natl. Cancer Inst. 13, 905-925 39 Ling, V. and Dixon, G. H. (1970) J. Biol. Chem. 245, 3035-3045 40 Lehmann, H. and Carrell, R. W. (1969) Br. Med. Bull. 25, 15-23 41 Fawcett, D. W., Anderson, W. A. and Phillips, D. M. (1971) Dev. Biol. 26, 229-251