Chemical Composition of the Acrosomes of Ram Spermatozoa

CHEMICAL COMPOSITION OF THE ACROSOMES OF RAM SPERMATOZOA

E. F. HARTREE and P. N. SRIVASTAVA A.R.C. Unit of Reproductive Physiology and Biochemistry, 307 Huntingdon Road, Cambridge {Received 27th May 1964)

Summary. Material extracted from ram spermatozoa with 0\m=.\0125 n-NaOH was separated into lipid and glycoprotein fractions. By use of an anionic detergent (Hyamine 2389) acrosomes were separated from ram spermatozoa and also fractionated into lipid and glycoprotein. The chemical compositions of the two glycoprotein fractions, as well as of the two lipid fractions, show marked similarities. Taking into consideration the chemical changes that may occur during the isolation of these fractions it is deduced that they provide a reasonable approximation to the composition of the acrosomes. Ofthe amino acids glutamic acid predom- inates. The following sugars are present: mannose, galactose, fucose, glucosamine, galactosamine and sialic acid. The residue left after extract- ing ram spermatozoa with n-NaOH contains polysaccharide of which the constituent sugars are glucose, galactose and mannose. All the sialic acid of ram spermatozoa appears to be contained in the acrosome. The sialic acid content of the spermatozoa of other species has been measured and the possible role of sialic acid in sperm physiology is discussed.

INTRODUCTION Clermont, Glegg & Leblond (1955) found that the component of the acrosome which is stained in the periodic acid-Schiff (pas) reaction could be removed from guinea-pig spermatozoa by extraction with 0-1 N-NaOH. Hathaway & Hartree (1963) showed that acrosomes can be removed from washed ram spermatozoa as effectively by 0-01 N-NaOH as by 0-1 N-NaOH; within that range of concentrations there was very little variation in the quantities of nitrogen and of orcinol-reactive sugar that passed into the alkaline extracts. This was taken as evidence that a discrete unit, the acrosome, was being dissolved. The possibility remained that material other than that present in the acrosome was simultaneously dissolved from the sperm cell in quantities which were also independent of the concentration of alkali. It was shown in the same paper that treatment of ram spermatozoa with hexadecyltrimethylammonium bromide (cetyltrimethylammonium bromide, ctab) caused membranous material, and in some cases apparently intact acrosomes, to be removed from the sperm head. Such detached elements, like the acrosomes of intact sperma¬ tozoa, were strongly stained by the pas and Giemsa techniques. 47

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access 48 E. F. Hartree and . . Srivastava We have now examined the effects ofother detergents upon ram spermatozoa. We have improved the methods for detachment of morphologically intact acrosomes and have been able to obtain them as suspensions which are virtually free from spermatozoa. The chemical composition of such acrosomal prepara¬ tions, and that of the material extracted from spermatozoa by alkali, has been determined. The sialic acid content of the semen of a few other species has also been measured.

MATERIALS AND METHODS Semen was collected twice weekly from a group of about twenty rams, and pooled. The spermatozoa were separated from the seminal plasma within 1 hr of collection and were washed with calcium-free Ringer solution as described by Hathaway & Hartree (1963). The final suspension was made up in Ringer solution to twice the original semen volume. Spermatozoa were stained with Giemsa solution as described by Hancock (1952) except that preparations were not fixed and the slides were left in the stain solution for 16 hr at 20° C. Bull and rabbit semen were dealt with in the same way. Epididymal bull semen was provided by Dr H. M. Dott and cock semen by Dr C. Polge.

DETACHMENT OF ACROSOMES BY DETERGENTS A suspension of washed ram spermatozoa in Ringer solution was centrifuged and the cells were made up to the same volume with 0-9% sodium chloride (saline). Portions of this suspension were mixed with equal volumes of solutions of the detergents in saline and were incubated at 37° C for 45 min. Observations made on stained preparations of spermatozoa treated by such methods are given in Table 1. Teepol XL and ctab removed the great majority of acro¬ somes, but both acrosomes and tails showed serious damage. Manoxol OT and Hyamine 2389 gave better results : the former was more effective in removing acrosomes but the latter caused less damage to the tails (PI. 1, Fig. 1). When the Hyamine-containing suspension was centrifuged for 15 min at 1500 g the super¬ natant fluid contained shrunken but characteristically stained acrosomes (PI. 1, Fig. 2). Further experiments with Hyamine led to a procedure by which 70 to 90% of acrosomes could be detached without appreciable damage to the tails. This is described below (B).

ISOLATION OF ACROSOMAL GLYCOPROTEIN FRACTIONS A. Isolation from an alkaline extract of spermatozoa A mixture of 40 ml of washed spermatozoa and 40 ml of 0-025 N-NaOH, having a pH of 11-4, was held at 37° C for 45 min and centrifuged at 6000g for 15 min. The almost clear supernatant fluid (75 ml) was brought to pH 6-5 with acetic acid and could be stored at —20° C if necessary. Each extract was mixed with four volumes of cold (—20° C) acetone, and centrifuged at 8000 g and 5° C for 10 min. The white sediment was suspended in about 10 ml of water and dialysed overnight at 5° C against 51. of distilled water. Lipid was removed from the dialysed solution by the procedure of Hartree & Mann (1961). The resulting precipitate was washed with chloroform and air-dried. This material

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access PLATE 1

Magnification of spermatozoa: x920 in Figs. 1 and 2; x2200 in Figs. 3 and 4. All preparations were stained by the Giemsa method. Fig. 1. Ram spermatozoa treated with 0-025% Hyamine for 45 min at 37° C. Fig. 2. Shrunken acrosomes separated from suspension shown in Fig. 1. Fig. 3. As Fig. 1, using 0-05% Hyamine for 90 min. Fig. 4. Acrosomes separated from suspension shown in Fig. 3.

(Facing p. 48)

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access Chemical composition of ram acrosomes

-a -o u 3 U S ( g •s 3|w S -S >s_bp ~3 "•¿ -a •S S c a S -a 5 c '> ta c o. OT3 " til 8" « ·ß.|c b s S s n u 1 ^ 5 ^ U d o « g a a-0 ¬ m3 ' ¡ OC S - 0 Su S « S c C u %r, sO ìn < ci rt SU tQ

»2 -û g o

<3* T3 rt S '3 ei ^_Q

s« s P "e 9

S3 o. J3 "5 "3 ¬ I «fe

ñ ü G M -à c c« M u c [ -t-» & 2- pu- .s « ¿--5^ •a-i?.? 3 S « S * ; í _. r! .a £ u 0 « 1 _, tZS iJJ O œ «- f-( "_ o a cm >,O.s .2* 3 V "o S m s -. o"S, -1 g-s 3 ai < Cß e H .rt t/j ss s 3 U

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access 50 E. F. Hartree and . . Srivastava will be referred to as glycoprotein A. The extracted lipid was obtained as a solution in chloroform-methanol. This was shaken with 0-2 vol of 0-01 M-MgCl2 to remove impurities and the lower layer was evaporated in a rotary evaporator. The lipid residue was stored in chloroform solution at —20° C (lipid A). The supernatant fluid from which the white sediment had been separated was concentrated to small bulk in a rotary evaporator and treated with an equal volume of 10% trichloroacetic acid. The solution remained clear and it was therefore assumed that all the protein had been precipitated by acetone.

-Washed ram spermatozoa

pH 11.4, Hyamine pH 6-1, centrifuge centrifuge I Supernatant Sedimenl Sediment solution (discarded)

resuspend, acetcne at centrifuge 5°C, centrifuge

Sediment Supernatant suspen¬ (discarded ) sion of acrosomes Supernatant Sediment solution Upoglycoprotein ethanci, (discarded) centrifuge chLoroform methanol - Supernatant Sediment Solution of Insoluble solution LIPID A residue (discarded) GLYCOPROTEIN A chloroform methanol

Solution of Insoluble LIPID residue GLYCOPROTEIN Flow sheets for separation of lipids and glycoproteins.

B. Isolation from detached acrosomes A suspension of washed ram spermatozoa (40 ml) in saline was incubated with an equal volume of 0-1% Hyamine, in 0-067 M-phosphate buffer pH 6-1, at 37° C for 90 min. At this stage some of the detached acrosomes were distorted but still clearly recognizable (PI. 1, Fig. 3). The suspension was centrifuged for 15 min at 1500g, the sediment was resuspended in fresh saline and centri¬ fuged again. The combined supernatant fluids contained acrosomes with very few (<1%) spermatozoa and sperm tails (PI. 1, Fig. 4). The sediment contained spermatozoa, of which about 80% were without acrosomes, and free acrosomes equivalent to about 20% of the spermatozoa. Thus the supernatant fluid con¬ tained in suspension approximately 60% of the acrosomes in the original sperm suspension. The supernatant fluid was treated with an equal volume of absolute ethanol, left overnight at 4° C, and centrifuged. The precipitate was resuspended in 7-5 ml ofwater and lipids were removed as described above. The final products were the air-dried glycoprotein and a solution oflipids in chloroform. These will

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access Chemical composition of ram acrosomes 51 be referred to as glycoprotein and lipid respectively. When the lipid solutions were required for analysis, phosphate buffer was replaced by saline as a solvent for Hyamine. This modification caused acrosomes to disintegrate more readily after detachment from the heads. A similar glycoprotein was obtained from bull spermatozoa by the same treatment.

ANALYSES Plasmalogen and acyl ester in the lipids were determined according to Hartree & Mann (1959). Residual acid-soluble and acid-insoluble phosphorus in the glycoprotein fractions were separated by heating the fractions with 5% trichloroacetic acid for 1 hr at 95° C and filtering. Total phosphorus was estim¬ ated by King's (1932) procedure; inorganic phosphorus was determined in the same way except that heating with perchloric acid was omitted. Neutral sugar in intact glycoproteins was determined by the orcinol method (Vasseur, 1948). To assay individual sugars, including amino sugars, the glyco¬ proteins were hydrolysed with a suspension of Dowex 50 X8 resin in 0-15 N-HC1 (Anastassiadis & Common, 1958; Hartree, 1964). By this treatment neutral sugars are obtained separately from amino sugars. Preliminary experi¬ ments showed that heating with resin for 30 hr at 100° C gave optimum release of sugars. The resulting solution of neutral sugars was brought to pH 5-5 with Amberlite IR-45, filtered and evaporated to dryness over KOH in a vacuum desiccator. The individual sugars were separated by descending paper chrom¬ atography, in ethyl acetate-pyridine-water (8 : 2 : 1 by vol), and eluted and analysed as described by Whistler & BeMiller (1962). Total amino sugars were analysed as described by Hartree (1964) and the ratio of galactosamine to glucosamine was measured by Garden's (1953) method. Results were cor¬ rected for losses incurred during resin hydrolysis of individual sugars. Burton's (1956) method for estimating DNA, but with deoxyribose standards, was used for determining bound deoxyribose in alkaline extracts of spermatozoa. Since only purine-bound deoxyribose reacts in this procedure (Lovtrup & Roos, 1963) analytical figures were doubled to give total bound deoxyribose. Sialic acid was released from semen, seminal plasma and glycoproteins by heating them for 1 hr at 80° C in 0-1 N-H2SO4. It was then determined by Aminoff's (1961) thiobarbituric acid method. To avoid interference from deoxyribose derivatives, spectrophotometric readings were made at two wave¬ lengths as described by Warren (1959a). Determination of sialic acid in sperm¬ atozoa was not possible by the above procedure because acid extracts contained too great a preponderance of deoxyribose derivatives. A preliminary fractionation removed the greater part of these derivatives and enabled an accurate assay to be obtained by the two-wavelength procedure. The assay was carried out as follows. The washed spermatozoa from 20 ml of semen were suspended in water to 90 ml, mixed with 10 ml of N-H2SO4 and heated for 1 hr at 80° C. The cooled mixture was neutralized to pH 6 with Ba(OH)2 and centrifuged, and the precipitate was washed twice with water. The combined clear fluids were passed through a column (10 x0-7 cm) of De-acidite FF, 200 mesh, (Permutit Co.) in the CL form and the column was washed with 10 ml

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access 52 E. F. Hartree and . . Srivastava of water. The combined effluents were evaporated in vacuum to 5 ml and analysed for sialic acid. Corrections were applied for the recovery of pure sialic acid from a similar column (88%). For amino acid analysis 30 mg of glycoprotein was heated in a sealed tube with 6 ml of 6 N-HC1 for 16 hr at 100° C. Hydrolysates were freed from HC1 by repeated evaporation with water. Aqueous solutions of the residues were passed through columns of Zeokarb 225, 200 mesh (Permutit Co.) in the H+ form. The columns were washed with water and amino acids were then eluted with 2 n- triethylamine in 20% acetone (Harris, Tigane & Hanes, 1961). Eluates were taken to dryness on a steam bath and amino acids in the residues were deter¬ mined with a Beckman automatic amino acid analyser. Glycoprotein from ram spermatozoa was examined by electrophoresis on paper at pH 9-5 and on starch gel at pH 8-1. (We are indebted to Dr J. P. Bennett for carrying out the latter experiment.) In both cases only one protein band was detected. Paper electrophoresis of glycoprotein A at pH 8-9 also gives rise to a single, but rather more diffuse band.

Table 2

comparison of lipids extracted from glycoproteins a (POOLED SAMPLES) AND (four SAMPLES) WITH LIPIDS OF WHOLE RAM SPERMATOZOA

m-equiv./g lipid Ratio Source of lipids Phosphorus ester to (%) Plasmalogen Acyl ester plasmalogen A 2-16 0-43 0-86 2-0 1 1-92 0-41 1-07 2-6 2 1-86 0-39 1-13 2-9 3 2-30 0-36 1-00 2-8 4 2-50 0-35 1-00 2-9 Spermatozoa* 2-5 0-47 1-26 2-7

Representative data based upon Lovern, Olley, Hartree & Mann (1957) and Hartree & Mann (1959, 1961).

RESULTS The material precipitated by acetone from an alkaline extract of ram sperm¬ atozoa and the acrosomal fraction obtained by treatment with Hyamine each contain about 20% lipid. Both bear some resemblance to the lipoglycoprotein complexes obtained from spermatozoa by Mayer {see Discussion).

COMPARISON OF LIPIDS A AND WITH THE LIPIDS OF RAM SPERMATOZOA The main feature of the lipids of ram spermatozoa is their high content of phospholipid, consisting mainly of the choline-based phosphatides plasmalogen and lecithin (Flartree & Mann, 1959, 1960, 1961). Table 2 shows analyses of the phosphorus, plasmalogen and acyl ester in the lipid fractions A and B, separated from the corresponding glycoproteins. The results obtained with lipid are close to those obtained previously with lipids of ram spermatozoa :

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access Chemical composition of ram acrosomes 53 in particular the ratio of ester to plasmalogen for lipid fractions is close to 2-7 which is characteristic of ram sperm lipids and indicates a plasmalogen/ lecithin ratio of 1-1 to 1-2 (Hartree & Mann, 1960, 1961). The lower ester/ plasmalogen ratio in lipid A is not unexpected since esters will be slowly hydrolysed under the conditions of alkaline extraction (pH 11-4).

PHOSPHORUS IN GLYCOPROTEINS A AND For phosphorus analyses (see Table 3) 150 mg samples of glycoproteins were suspended in 5 ml of water and dialysed against three 150 ml portions of 5 mM-HCl. The diffusâtes were evaporated to dryness and the residues dissolved in a little water. The phosphorus in these fractions consisted mainly or entirely of inorganic phosphate deriving from the Ringer solution : other Ringer anions were present only in traces. The non-diffusible glycoprotein fractions remaining in the dialysis tubing contained small amounts of phosphorus which are tentatively ascribed to residual DNA and lipid (Table 5).

Table 3

distribution of phosphorus between diffusible and non- diffusible material of glycoprotein fractions a and ß

Original glycoprotein Diffusible fraction Non-diffusible fraction acid-soluble 0-026 0-11 acid-insoluble 0-033 0-084

Results for inorganic phosphorus are given in brackets : all other values are total phosphorus (see also Table 5).

SUGARS AND AMINO ACIDS IN GLYCOPROTEINS A AND Air-dried glycoprotein A contained 79-6% non-diffusible solids, 10-6% diffus¬ ible (i.e. dialysable) solids (mainly inorganic phosphate) and 9-8% water. The corresponding figures for were 76-7, 10-2 and 13-1 respectively. After acid hydrolysis the diffusible fractions contained negligible amounts of amino acids and sugars. The analyses described in this section were carried out on air-dried glycoproteins and corrections were applied to obtain the compositions of the non-diffusible (i.e. salt-free) solids. Detailed analytical results are given in Tables 4 and 5. A complete recovery of amino acids was not obtained : values for cysteine, methionine and tyrosine can be assumed to be low because of decomposition during hydrolysis. Molar ratios for sugars are based upon the assumption that the most accurate analysis (glucosamine+galactosamine) represents a whole number, i.e. 9, which is consistent with the presence of one molecule of fucose. Determinations of neutral sugars are liable to errors up to 10%.

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access 54 E. F. Hartree and . . Srivastava Table 4

NEUTRAL SUGARS, AMINO SUGARS, SIALIC ACID AND AMINO ACIDS IN GLYCOPROTEINS A AND

vmolesj: mg/g Molar ratios (anhydroform)

Fucose 14 19 2-1 2-8 0-9 1-0 Galactose 62 71 10 12 3-9 3-7 Mannose 93 121 15 20 5-9 6-3 Galactosamine 45 53 7-2 8-6 2-8 2-8 Glucosamine 98 119 15-7 19-1 6-2 6-2 Sialic acid 14 37 4-0 10-6 0-9 1-9 Lysine 485 455 61-9 58-3 2-9 3-1 Histidine 159 144 21-8 19-8 1-0 1-0 Ammonia 237 148 1-4 1-0 Arginine 330 290 51-5 45-1 2-0 2-0 Aspartic acid 822 736 94-6 85-0 5-0 5-0 Threonine 498 448 50-5 43-3 3-0 3-1 Serine 611 560 53-0 49-0 3-7 3-8 Glutamic acid 980 720 126-3 93-0 5-9 4-9 Proline 575 405 55-7 39-4 3-5 2-8 Glycine 605 545 34-5 31-3 3-7 3-7 Alanine 615 535 43-7 37-9 3-7 3-7 J Cystine 66 84 6-8 8-7 0-4 0-6 Valine 472 429 46-8 42-5 2-9 2-9 Methionine 166 138 21-7 18-2 1-0 1-0 Isoleucine 346 314 39-1 35-5 2-1 2-1 Leucine 743 614 84-2 69-1 4-5 4-2 Tyrosine 236 228 38-6 37-2 1-4 1-6 Phenylalanine 304 295 44-6 43-3 1-8 2-0

Table 5

summary of compositions of glycoproteins A AND

Amino acids Sugars Phospholipid (25 acid-insoluble P) Nucleic acid (10 acid-soluble P)

* Attempts to confirm the DNA content by Burton's (1956) method were not successful owing to interference by other sugar components. However, it was clear that not all the acid-soluble could be accounted for as DNA.

CARBOHYDRATE IN THE ALKALI-INSOLUBLE PORTION OF RAM SPERMATOZOA According to Hathaway & Hartree (1963) the amount of orcinol-reactive sugar that is extractable from ram spermatozoa by NaOH at concentrations between 0-01 and 0-1 is more or less constant. However, with higher concentrations

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access Chemical composition of ram acrosomes 55 ofNaOH the orcinol reaction in the extracts becomes more intense until at 1-0 it represents approximately twice the constant amount found with lower concentrations (top curve of Text-fig. 1 ). The bottom curve in this figure is the calculated contribution to the orcinol colour reaction of the deoxyribose present in the extracts as DNA. This contribution was determined from assays of nucleotide deoxyribose by Burton's (1956) method and from the finding that 1 µ of deoxyribose gives the same colour in the orcinol reaction as 0-19 µ ß of glucose. The difference curve (pecked line) thus records the amount of non-DNA, orcinol-reactive sugar that is extracted by different concentrations of NaOH. This amount is almost constant within the range 0-01 to 1-0 n-

10 25 100 250 1000 t mM -NaOH in extraction medium Whole spermatozoa Text-fig. 1. Extraction of carbohydrate from washed ram spermatozoa by different concentrations of NaOH. Suspensions ofspermatozoa in Ringer solution were incubated with equal volumes of NaOH solution for 45 min at 37° C and centrifuged. D, Orcinol-reactive carbohydrate estimated as glucose. O, Values for bound deoxyribose determined by Burton's (1956) method and converted to their glucose equivalents in the orcinol reaction as described in the text. The pecked line is a difference curve.

NaOH. Analytical figures for whole spermatozoa were obtained by analysing the suspension of spermatozoa in N-NaOH before removal of the insoluble residue. Text-fig. 1 shows that this residue contains carbohydrate but no deoxyribose. A sample of the alkali-insoluble residue was washed successively in a centrifuge with N-NaOH (twice), water and ethanol. After the product had been hydrolysed by the resin method and examined by paper chromato¬ graphy, mannose, galactose and glucose were detected, the last-named pre¬ dominating. Hexosamines were virtually absent.

LOCALIZATION OF SIALIC ACID IN RAM SPERMATOZOA From an analysis of washed spermatozoa it was calculated that the spermatozoa present in 100 ml of ram semen contain 2-61 mg sialic acid (the corresponding value for ram seminal plasma is 50 to 60 mg). Treatment of glycoprotein with

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access 56 E. F. Hartree and . . Srivastava 0-1 N-H2SO4, followed by direct estimation of sialic acid showed that this material contained 0-59% sialic acid. From 100 ml of ram semen the average yield of is 287 mg and since recovery of acrosomes is about 60% the total acrosomal material would be about 480 mg per 100 ml of semen. This would contain 480x0-0059 or 2-84 mg of sialic acid. The agreement with 2-61 is within experimental error and it can therefore be concluded that all the sialic acid of ram spermatozoa is in the acrosomes.

SIALIC ACID IN SEMEN OF OTHER SPECIES Warren (1959b) has published analyses of human semen for sialic acid. These are included in Table 6 which summarizes our findings with semen from bulls, rams, rabbits and cocks. In all cases the sialic acid is present mainly in the plasma. No free sialic acid has been detected in fresh material. In the case of washed ram spermatozoa it seems unlikely that the sialic acid is contributed by Table 6

sialic acid in semen

Bull Human Ram Rabbit Cock Hereford Red Poll Friesian Seminal plasma 402* 317 293 405 178f 72 40 (208 to 675) (156 to 209) Spermatozoa 2* 6-6 4-0 7-0 8-5 9-3

Results are expressed as nmoles in material from 100 ml of semen. * Warren (1959b). Values entered under seminal plasma are those for whole semen (twenty samples). The correction for spermatozoal sialic acid would be within experimental error, t Mean of four samples each consisting of twenty pooled ejaculates. X Not detectable in 7 108 spermatozoa. residual plasma : thus the sialic acid in such spermatozoa is recovered quantita¬ tively in glycoprotein B. Warren detected sialic acid in human spermatozoa after very extensive washing. The figures for bull spermatozoa in Table 6 are comparable with those of ram spermatozoa, but since the sperm density of bull semen is, on average, only one third that of ram semen (Mann, 1954), the amount of sialic acid per spermatozoon, and consequently the concentration of sialic acid in glycoprotein B, should be two to three times higher for bull than for ram. However, one sample of bull glycoprotein contained only 0-07% sialic acid compared with 0-59% in ram glycoprotein B. Thus consider¬ able loss of sialic acid appears to have occurred during isolation of the bull glycoprotein. Unwashed epididymal spermatozoa from a Friesian bull con¬ tained 0-08 µ ß of sialic acid per 109 cells which is not significantly different from the value 0-07 for washed ejaculated spermatozoa. DISCUSSION

COMPOSITION OF THE RAM SPERM ACROSOME Treatment of ram spermatozoa with dilute NaOH leads to dissolution of the acrosome or at least of that part of the acrosome that takes up Giemsa stain

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access Chemical composition of ram acrosomes 57 (Hathaway & Hartree, 1963). However, the alkaline extract could well contain material extracted from other parts of the spermatozoon. On the other hand, while treatment with detergents leads to detachment of apparently intact acrosomes, the separated acrosomes subsequently become distorted and shrunken and it cannot be assumed that no material has been lost from them. However, the chemical compositions of glycoproteins A and B, which were isolated from ram spermatozoa by two quite distinct methods, show consider¬ able similarities. The most reasonable conclusion from this is that the solubiliz- ing action of alkali is limited to the acrosome and that the effect ofdetergent is to detach the acrosome but not to solubilize any constituent of it. On the other hand such similarities would still be found if any non-acrosomal material which may dissolve in the alkaline extraction medium, or any acrosomal component which may be solubilized by detergent, has a chemical composition close to that of the whole acrosome. In either case it follows that the probable composi¬ tion of the defatted acrosome is intermediate between the analyses given for glycoproteins A and B. Certain differences between A and must be considered. The significant differences between the total sugar contents of the two fractions (Table 5) may be due to alkali-sensitive linkages between the polysaccharide (or its oligo- saccharide components) and the protein (Gottschalk & Graham, 1959; Cook, 1962). The lower sialic acid content of glycoprotein A is also consistent with the action of an alkaline extraction medium. In addition there is evidence that alkali removes some non-acrosomal material from ram spermatozoa (Hathaway & Hartree, 1963). We therefore believe that the analyses of glycoprotein are the more reliable guide to the composition of the non-lipid portion of the ram acrosome. Another significant difference between glycoproteins A and B, for which we can offer no explanation at present, is in the levels of glutamic acid. The acrosomal phospholipid appears to have a composition similar to that of the whole spermatozoon. However, it is not possible to rule out the possibility that treatments with alkali and with detergent may lead to association of acrosomal protein with lipids derived from other parts of the sperm cell. Mayer and his associates (Thomas & Mayer, 1949; Mayer, 1955; Miller & Mayer, 1960) have studied a lipoglycoprotein fraction which is removed by alkali from bull and boar spermatozoa. In this material they found cholesterol, phospholipids, protein, glucosamine and glucuronic acid. In our experiments a wider range of sugars was detected but the quantities of uronic acid in glyco¬ proteins A and from ram spermatozoa were too low for assay. However, Mayer washed spermatozoa with water, freeze-dried them, and extracted them with alkali in a tissue grinder; treatments which may influence the nature of the material extractable by alkali. Miller & Mayer (1960) reported that the alkali-soluble material from bull spermatozoa appears homogeneous on electro¬ phoresis : our glycoproteins A and from ram spermatozoa behave similarly. For the preparation of glycoprotein B, ram spermatozoa are superior to bull spermatozoa. In the presence of Hyamine, or of other detergents, bull acrosomes tend to break up readily : if the pH is below neutrality clumping of acrosomal fragments and agglutination of spermatozoa occur, while in more alkaline solutions the acrosomes disintegrate further.

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access 58 E. F. Hartree and . . Srivastava

ALKALI-RESISTANT RESIDUES OF SPERMATOZOA When Mayer and his co-workers had extracted the lipoglycoprotein from mammalian spermatozoa, they obtained an insoluble residue which contained DNA and an arginine-rich protein. When this material was extracted more exhaustively with alkali, the final residue contained 'ghost' sperm heads and fine granular material. Green (1940) subjected ram spermatozoa to prolonged extractions with dilute alkali and dilute acid and obtained an arginine-rich protein that was free from nuclear material. However, in neither case was the presence of sugar reported. We find that treatment of washed ram spermatozoa with N-NaOH for 45 min at 37° C leads to almost complete solubilization of the sperm cell. The insoluble residue contains polysaccharide which is distinct, in terms of solubility, chemical composition and intracellular location, from the acrosomal glyco¬ protein. Text-fig. 1 suggests that this polysaccharide may represent about 30% of the total polysaccharide. Even if this figure is only approximately correct it follows that the pas staining reaction is an unsatisfactory guide to the quantita¬ tive distribution of polysaccharide in spermatozoa.

SIALIC ACID AND THE CHARGE ON THE SPERMATOZOON Suspensions ofliving cells in media of physiological ionic strength and pH have a negative surface charge which is made apparent by their movement in a potential gradient. The charge on the erythrocyte, and on some other cells, is markedly reduced when the cells are treated with neuraminidase (see sum¬ maries by Gottschalk, 1954; Eylar, Madoff, Brody & Oncley, 1962). Neur¬ aminidase specifically releases sialic acid from its linkages, in glycoproteins, with either galactose or jV-acetylgalactosamine (Gottschalk, 1957, 1960) and it is clear that the negative charge on the erythrocyte is due largely to ionized sialic acid residues. Mammalian spermatozoa also exhibit a negative charge (Nevo, Michaeli & Schindler, 1961) which is considerably reduced when they are treated with neuraminidase (Fuhrmann, Granzer, Bey & Ruhenstroth-Bauer, 1963). Bedford (1963) has found that the negative charge on rabbit spermatozoa decreases as they move through the epididymis from the caput to the corpus. Fuhrmann et al. (1963) also noted that epididymal bull spermatozoa carry a higher negative charge than the spermatozoa in whole semen. This charge rose when spermatozoa were separated from seminal plasma and washed, but fell again when the washed cells were transferred to plasma. Our analyses of bull spermatozoa suggest that these changes are due to absorption of positively charged components of accessory gland secretions rather than to changes in the sialic acid content of the spermatozoa. According to Nevo et al. (1961) the tails of bull spermatozoa carry a higher negative charge density than the heads. Until the distribution of sialic acid in bull spermatozoa has been determined a clear interpretation of these results is not possible. However, the observed loss of sialic acid when glycoprotein is prepared from bull spermatozoa (see above) is consistent with the presence of non-acrosomal sialic acid. On the basis of their electrophoretic experiments Fuhrmann et al. (1963) calculated that the

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access Chemical composition of ram acrosomes 59 sialic content of bull spermatozoa could vary between 3-8 and 9-6 IO-8 µg per cell. Our estimate, based upon the mean sperm density of semen from a Friesian bull (Table 6), is 4-3 X 10~8 µg while Warren (1959b) gave the value 6·5 X IO-8 µg for the human sperm cell. The hyaluronic acid in the corona cells of the unfertilized ovum con¬ tributes a negative charge that will exert a repellent effect upon a negatively charged sperm cell. Factors which would tend to nullify this effect are sperm motility and reductions in the charge densities. It has been pointed out above that the negative charge on the spermatozoon may become reduced in vivo by adsorption of positive ions. Our finding that neuraminidase can be extracted from the mucosa of the rabbit uterus raises the possibility that reduction of charge on the spermatozoon may also occur in the uterus by release ofsialic acid from sperm glycoproteins. ACKNOWLEDGMENTS

We wish to thank Dr . M. Dott for his critical reports on our stained prepara¬ tions, Dr J. I. Harris and Mr R. E. Offoerd for carrying out the amino acid analyses, and Dr J. Burton and Dr P. A. Anastassiadis for advice on analytical procedures. Our thanks are due also to Dr T. R. R. Mann, f.r.s., for his interest in this work and for numerous stimulating discussions. One of us (P.N.S.) is grateful to the Colombo Plan authorities for the award of a fellowship.

REFERENCES Aminoff, D. (1961) Methods for the quantitative estimation of JV-acetylneuraminic acid and their application to hydrolysates of sialomucoids. Biochem. J. 81, 384. Anastassiadis, P. A. & Common, R. H. (1958) Liberation ofhexosamine, hexuronic acid, and hydroxy¬ proline from tissues by resin hydrolysis. Canad. J. Biochem. 36, 413. Bedford, J. M. (1963) Changes in the electrophoretic properties of rabbit spermatozoa during passage through the epididymis. Nature, Lond. 200, 1178. Burton, K. (1956) A study of the conditions and mechanism of the diphenylamine reaction for the colorimetrie estimation of deoxyribonucleic acid. Biochem. J. 62, 315. Clermont, Y., Glegg, R. E. & Leblond, C. P. (1955) Presence of carbohydrate in the acrosome of guinea-pig spermatozoa. Exp. Cell Res. 8, 453. Cook, G. M. W. (1962) Linkage of sialic acid in the human erythrocyte ultra-structure. Nature, Lond. 195, 159. Eylar, E. H., Madoff, . ., Brody, O. V. & Oncley, J. L. (1962) The contribution of sialic acid to the surface charge of the erythrocyte. ,7- biol. Chem. 237, 1992. Fuhrmann, G. F., Granzer, E., Bey, E. & Ruhenstroth-Bauer, G. (1963) Die elektrophoretische Beweglichkeit der Spermien des Rindes. - Naturforsch. 18b, 236. Gardell, S. (1953) Separation on Dowex 50 ion exchange resin ofglucosamine and galactosamine and their quantitative determination. Acta chem. scand. 7, 207. Gottschalk, A. (1954) The influenza virus enzyme and its mucoprotein substrate. Tale J. Biol. Med. 26, 352. Gottschalk, A. (1957) The structure of the prosthetic group of bovine submaxillary gland muco¬ protein. Biochim. biophys. Acta, 24, 649. Gottschalk, . (1960) Correlation between composition, structure, shape and function of a salivary mucoprotein. Nature, Lond. 186, 949. Gottschalk, A. & Graham, E. R. B. (1959) 6-a-D-sialyl-./V-acetylgalactosamine: the neuraminidase- susceptible prosthetic group of bovine submaxillary mucoprotein. Biochim. biophys. Acta, 34, 380. Green, W. W. (1940) The chemistry and cytology of the sperm membrane of the sheep. Anat. Ree. 76, 455. Hancock, J. L. (1952) The morphology of bull spermatozoa. J. exp. Biol. 29, 445. Harris, C. K., Tigane, E. & Hanes, C. S. (1961) Quantitative Chromatographie methods: isolation of amino acids from serum and other fluids. Canad. J. Biochem. 39, 439.

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access 60 E. F. Hartree and . . Srivastava Hartree, E. F. (1964) Observations on the determination of hexosamine in hydrolyzates of glyco¬ proteins. Analyt. Biochem. 7, 103. Hartree, E. F. & Mann, T. (1959) Plasmalogen in ram semen, and its role in sperm metabolism. Biochem. J. 71, 423. Hartree, E. F. & Mann, T. ( 1960) Phospholipids in mammalian semen. J. Reprod. Fértil. 1,23. Hartree, E. F. & Mann, T. (1961) Phospholipids in ram semen: metabolism ofplasmalogen and fatty acids. Biochem. J. 80, 464. Hathaway, R. R. & Hartree, E. F. (1963) Observations on the mammalian acrosome: experimental removal of acrosomes from ram and bull spermatozoa. J. Reprod. Fértil. 5, 225. King, E. J. (1932) The colorimetrie determination ofphosphorus. Biochem. J. 26, 292. Lovern, J. ., Olley, J., Hartree, E. F. & Mann, T. (1957) The lipids of ram spermatozoa. Biochem. J. 67, 630. Lovtrup, S. & Roos, K. (1963) Further observations on the chemical determination of deoxyribonucleic acid in animal tissues. Biochim. biophys. Acta, 68, 425. Mann, T. (1954) The biochemistry of semen. Methuen, London. Mayer, D. T. (1955) Reproduction and infertility. Michigan State University Centennial Symposium Report. Michigan State University, East Lansing, U.S.A. Miller, L. D. & Mayer, D. T. (1960) Survival of bovine spermatozoa in media containing a lipo- protein complex isolated from spermatozoa and other sources. Res. Bull. Mo. Agrie. Exp. Sta. No. 742. Nevo, A. C, Michaeli, I. & Schindler, . (1961) Electrophoretic properties of bull and rabbit spermatozoa. Exp. Cell Res. 23, 69. Thomas, L. E. & Mayer, D. T. (1949) The proteins of mammalian spermatozoa and cellular nuclei. Science, 110, 393. Vasseur, E. (1948) A spectrophotometric study of the orcinol reaction with carbohydrates. Acta chem. scand. 2, 693 Warren, L. (1959a) The thiobarbituric acid assay of sialic acids. J. biol. Chem. 234, 1917. Warren, L. (1959b) Sialic acid in human semen and in the male genital tract. J. clin. Invest. 38, 755. Whistler, R. L. & BeMiller, J. N. (1962) Methods in carbohydrate chemistry, vol. 1. Eds. R. L. Whistler & M. L. Wolfram. Academic Press, New York.

Downloaded from Bioscientifica.com at 09/30/2021 08:48:50AM via free access