The Biosynthesis of Protamine in Trout Testis

SUMMARY Protamine isolated from rainbow trout testis can be separated into three main components, Cr, Crr, and Crrr, by chromatography on microgranular carboxymethyl cellulose columns. After further purification by rechromatography on the same column, each component was analyzed for amino acid content. These analyses indicated that the components were distinct but closely related. However, each component still appeared heterogeneous since the number of residues of different amino acids assumed nonintegral values. The relative amounts of the protamine components present in spermatid nuclei were different at different stages of spermatogenesis. It appeared that the relative amount of Cr decreased while C rrr increased during testis maturation probably as the result of differing rates of synthesis of each component. The probable function of protamine is in the packaging of DNA into the sperm head. The independent synthesis of protamine components suggests that, in spite of their similarities, they may have different specific functions.

show that disomes are labeled more rapidly than heavier polysomes. The distribution of incorporated 14C-arginine and r4C-lysine through the polysome profiles changes markedly during the maturation of the testis; at the time of rapid protamine synthesis, the disomes incorporate arginine rapidly but lysine slowly, suggesting that they are largely engaged in protamine synthesis. Actinomycin D has an inhibitory effect on the incorporation of 14C-arginine into protamine only during the early protamine stage of development, suggesting that messenger RNA for protamine is synthesized before protamine synthesis takes place and is metabolically stable.
We have previously presented several lines of evidence that at a late stage of spermatogenesis protamine is synthesized in the cytoplasm of rainbow trout testis cells and is quickly transported into the nucleus where it eventually replaces the histones on DNA (1). A special class of cytoplasmic polysomes, the di- ribosomes (disomes), sedimenting at 120 S, appears highly active in incorporating 14C-arginine and has been implicated as the species of polysome synthesizing protamine (1). In this present report, we substantiate further our previous observation of cytoplasmic disomes as the site of protamine synthesis and present evidence of specific developmental changes in the testis polysome population as the testis cell matures. Concomitant with the changes in polysome profiles, there are major alterations in the relative incorporation and distribution of labeled lysine and arginine in the polysomes which suggest an ordered synthesis of different proteins as development proceeds. MATERIALS AND METHODS Trout TestisTestes at various stages of sexual maturation were obtained at the middle of each month during August to December 1968 from naturally maturing rainbow trout (S&no guirdnerii) (13 to 18 months old) raised in the Sun Valley Trout Farm, Coquitlam, British Columbia. Excised testes were placed on ice, transported to the laboratory, and washed with cold running tap water.
A number of testes (usually about 20) were pooled and minced with scissors, and a portion of the mince was removed for incubation with radioactive amino acids. The excised testis could be kept on ice for up to 4 hours without any appreciable loss in ability to incorporate labeled amino acids into protamine.
The testes not required for incubation were quickly frozen on Dry Ice and stored at -80" until use.
Cell Incubations-Cell suspensions were prepared with 10 to 20 g of fresh testes in Buffer TMKS (Tris-HCl, 0.05 M; magnesium acetate, 0.005 M; potassium chloride, 0.025 M; and sucrose, 0.25 M), pH 7.6, as previously described (1). Identical aliquots of the cell suspension were incubated separately with 14C-arginine and i4C-lysine (0.5 &i per ml of labeled amino acid per incubation mixture, specific activities of 226 to 241 mCi per mM, New England Nuclear) at 20" for 30 min in a gyratory water bath (New Brunswick).
Incubations were terminated by the addition of cycloheximide (Calbiochem) to a final concentration of 0.4 MM.
Isolation of Ribosomes-Ribosomes were prepared from trout testis and liver by a met.hod similar to that described by Wettstein, Staehelin, and No11 (2). All operations were performed at 2-4".
Frozen or fresh tissue was minced with scissors in 1 to 2 volumes of Buffer TMKS and homogenized in a Tri-R tissue homogenizer (Tri-R Instruments, New York) at 5000 rpm for 20 sec. The homogenate was centrifuged at 15,000 x g for 15 min to obtain a postmitochondrial supernatant.
The post  Centrifugation was performed at 60,000 rpm for 2 hours in a No. 65 rotor (Spinco), the supernatant above the dense sucrose layer was aspirated off, and then the dense sucrose layer itself was similarly removed, leaving a clear ribosomal pellet.
The inside wall of the centrifuge tube was wiped with absorbent paper tissue and the ribosomal pellet was briefly rinsed with a small volume of cold Buffer TMK.
The pellet was then dissolved in Buffer TMK, yielding an opalescent solution.

Sucrose Density Gradient
Analysis of Ribosomes-Sucrose solutions of appropriate concentrations were prepared in Buffer TMK with ribonuclease-free density gradient grade sucrose (Mann). Linear sucrose gradients (35 ml) of 10 to 34% w/v were generated by a Beckman density gradient former (DGF-IM- 3) in cellulose nitrate tubes at room temperature.
The gradients were then cooled at 4' for 2 to 3 hours. Appropriate concentrations of ribosome solutions were made up in Buffer TMK, layered onto the gradients, and centrifuged at 4" in a SW-27 swinging bucket rotor (Spinco) at 26,000 rpm for 2.5 hours. The rotor was allowed to stop without braking.
The centrifuge tube was punctured at the bottom with a needle and equal fractions were collected for analysis of absorbance at 260 nm and radioactivity. For the assay of radioactivity, an aliquot was removed from each fraction, carrier proteins (bovine serum albumin and protamine) were added, and total protein was precipitated with cold pH 2 trichloracetic acid-tungstate solution (3). The precipitate was collected by centrifugation, washed once with trichloracetic acidtungstate, resuspended in 0.5 ml of distilled water, and transferred to a glass scintillation vial. Radioactivity was assayed in 10 ml of Bray's fluid (4) with a Unilux liquid scintillation counter.
For a more sensitive analysis of the ribosome profile, the effluent obtained from the sucrose gradient after puncturing the centrifuge tube was continuously monitored and recorded at 254 nm with an Isco ultraviolet analyzer equipped with a IO-mm flow cell. The flow rate of the effluent was kept constant by means of an Accu-flow pump (Beckman). Incorporation of I%-Arginine into Protamine in Presence of Actinomycin D-Two testes (3 g each) were obtained from a naturally maturing trout at the early protamine stage of spermatogenesis (Table I). A cell suspension prepared from the testes was divided into two equa.1 portions; one was kept as control and to the other was added actinomycin D to a final concentration of 0.01 mM. After a preliminary incubation at 20" for 30 min, 1 PCi of 14C-arginine was added to each suspension and the incubation was continued for 60 min, after which cycloheximide was added to a final concentration of 0.4 mM to terminate incorporation. Basic proteins were extracted from the cell nuclei of each suspension and protamine was isolated by chromatography on a Bio-Gel P-10 column as previously described (1). The total amount of radioactivity associated with protamine was determined for each suspension.

Trout Testis Development and Time of Protamine Xynthesis-
The immature testis collected from 1.5-to 2-year-old rainbow trout during August weighs approximately 1.6 g and rapidly increases in weight to a maximum of 10 g by October (Table I). During this period of rapid weight increase, the spermatogonial stem cells in the testis are presumably undergoing repeated mitoses to form a large number of cells which, after the meiotic divisions, subsequently develop into sperm cells (6,7). By December, however, there is a sharp decrease in the average weight of the testis, partly as a result of the release of mature spermatozoa.
These fish normally spawn during January and February.
The rate of incorporation of i4C-arginine into protamine varies considerably as development progresses (Table I). As previously shown (l), at the stage when protamine synthesis is rapid there is little histone synthesis. Although the differentiation of secondary spermatocytes first to spermatids and then to mature spermatozoa is occurring during this period, no cell divisions are involved and there is no increase in cell number.
When the greater proportion of the cells in the testis is mature spermatids and sperm cells, protamine synthesis has largely ceased (Table I). From Table I  Ribosomes were isolated from testis and liver obtained from naturally maturing trout at the middle of each month as described under "Materials and Methods." Purified ribosomes were analyzed by sucrose density gradient (10 to 3470 w/v, 35 ml) centrifugation at 26,000 rpm. for 2.5 hours in a Spinco SW-27 swinging bucket rotor. The bottom of the centrifuge tube was punctured and the polysome profile in the efRuent was monitored on the Isco untraviolet analyzer as described under "Materials and Methods." late protamine stage in December when there is an even sharper decrease. This decrease in the number of ribosomes during spermatogenesis is consistent with the observations of various investigators (9)(10)(11) who have shown that cytoplasm is sloughed off from developing sperm cells in rat testis.
Creelman and Tomlinson (12) have also observed a progressive decrease in RNA content in the maturing salmon testis.
It is clear also that the total number of ribosomes present at a particular stage does not reflect the rate of 14C-arginine incorporation into protamine since, at the preprotamine stage when little protamine is synthesized, there are more ribosomes than during the protamine stage when protamine is rapidly synthesized.
Polysome Projiles during profiles of trout testis were examined at various stages of spermatogenesis, changes in the relative amounts of different classes of polysomes were observed (Fig. 1). The monosomes and the disomes sediment at 77 S and 120 S, respectively (1). It may be seen in Fig. 1 that large peaks of disomes are present in the polysome profiles obtained from October and November testes. In September and December, these disomes are present in only relatively small numbers. It is significant that protamine synthesis is most rapid during October and November when the relative proportion of disomes in the polysome population is greatest (Table I). This suggests that the rate of protamine synthesis is a function of the number of disomes present and is consistent with previous observations (1) that protamine is synthesized on disomes. During late November, the number of monosomes increases sharply, although at this stage the disome peak is still large. The source of these monosomes appears to be a degradation of the larger polysomes but in December, when there is a marked decrease in the total number of ribosomes (Table I), the disomes themselves largely disappear and the great majority of polysomes are converted to the monomeric state.
Effect of Mg++ on Polysome ProfilesVarious investigators (13)(14)(15) have noted that dimeric ribosomes sedimenting at 110 to 120 S are formed as the result of an artifactual aggregation of ribosomes at high magnesium concentrations.
Since these dimeric ribosomes either dissociate at 1 mu Mg* ion concentration or are absent when extracted at this lower magnesium concentration (15), two parallel extractions of polysomes were performed on testes obtained at the protamine stage. Two TMK buffers, one containing the usual 5 mM magnesium acetate and the other 1 mM magnesium acetate, were used for preparation of polysomes.
It may be seen in Fig. 2 that polysomes extracted with 5 mM Mg++ show few monosomes.
Conversely, the ribosomes extracted with 1 mM Mg* appear to have substantially fewer large polysomes than those extracted with 5 mM Mg++. However, the two profiles are similar in that both have prominent peaks of disomes sedimenting at 120 S. Since the disomes are Several other lines of evidence support this conclusion. Reader and Stanners (15) have observed that the ability to form Mg++-dependent dimeric ribosomes varies with different species of animals but ribosomes extracted from the various tissues of the same animal form dimeric ribosomes to the same extent.
In the case of the rainbow trout, prominent peaks of disomes are observed only in ribosomes isolated at a specific time, namely the protamine stage of spermatogenesis, and, further, liver ribosomes extracted from the same fish, under the same conditions, during the protamine stage of testis development, do not show this large peak of disomes (see Fig. I). In addition, Mg++dependent dimeric ribosomes are usually inactive in protein synthesis (14,15), having little nascent protein associated with them, but, as has been previously shown (I), trout testis disomes are highly active in synthesis and contain most of the nascent protamine.

Effect of Pancreatic
Ribonuclease on Polysome Profiles--The decrease in the proportion of larger polysomes observed in ribosomes extracted in 1 mM Mg++ (Fig. 2B) is likely to be the result of degradation of these polysomes by nucleases released during isolation in the presence of the lower magnesium ion concentration.
This, however, would suggest that the disomes observed in Fig. 2 are resistant to nuclease action since there appears to be no decrease in the number of disomes. The effect on testis polysomes of a mild treatment with pancreatic ribonuclease was therefore examined (Fig. 3). The ribosomes were purified from testis cells previously incubated with 14Carginine and hence contained labeled nascent proteins associated with them. It may be seen in Fig. 3A that there are two major peaks of 260 nm absorbance (&l and D) in the polysome profiles corresponding to monosomes and disomes, respectively. The major peak of radioactivity is associated with the disomes (D) and radioactivity is also found in the large polysomes. After ribonuclease treatment (Fig. 3C), there is a decrease in the amount of absorbance at 260 nm in the large polysome region and an increase in monosomes (M).
Radioactivity associated with the larger polysomes is also removed by the ribonuclease treatment.
When bentonite (19, 20) is added before the ribonuclease to moderate the severity of the treatment (Fig. 3B), there is partial protection of the larger polysomes and some 14Carginine remains associated with them. These observations are consistent with those of other investigators (16,18) who have studied the effect of mild ribonuclease treatment on bacterial and animal polysomes.
As may be seen from Fig. 3C, the disomes (D), in contrast to the large polysomes, are relatively unaffected by mild ribonuclease treatment.
Time Dependence of Incorporation into Polysomes and Disomes -To determine whether the disomes might result from degradation of larger polysomes during the isolation procedure, ribosomes were prepared from two identical testis cell suspensions after a pulse of 14C-arginine at 20" for 2 min and 40 min, respectively. The polysome profile in cells from each incubation was examined by sucrose density gradient centrifugation (Fig. 4). After 2 min of incorporation, the major peak of radioactivity is associated with disomes sedimenting at 120 S while the larger polysomes and the monosomes (77 S) contain little radioactivity.
After 40 min of incorporation, however, the amount of radioactivity associated with the disomes and monosomes remained essentially the same as that observed after 2 min of incorporation, but now more radioactivity is associated with the larger polysomes. Since the 14C-arginine appears first on the disomes and only after a longer incubation period on the larger polysomes, the disomes cannot be the breakdown products of larger polysomes.
Further, nuclease usually degrades polysomes into monosomes (16)(17)(18) and since after a longer incubation period the increase in label on the larger polysomes is not reflected by an increase in radioactivity on the monosomes (Fig. 4), it appears that testis polysomes are degraded neither by nuclease during the incubation nor during the isolation procedure at the standard 5 mM Mg++ concentration.
Mitochondrial Ribosomes-In the treatment of the postmitochondrial cytoplasmic fraction with Triton X-100 during the preparation of ribosomes, there was the possibility that a small number of mitochondria still present after differential centrifugation would be lysed (21)  would then be released into the cytoplasm. As a result, it was possible that the peak of radioactive protamine appearing in the disome region of the cytoplasmic testis polysomes (1) could be due to the presence of mitochondrial ribosomes.
To check the possibility of a mitochondrial site of synthesis for protamine, the ribosomes of the postnuclear supernatant (cytoplasm plus mitochondria) and those of the postmitochondrial supernatant were analyzed after testis cells were incubated with K-arginine for 2 min at 20". The results are shown in Fig. 5, A, B, and C. The "mitochondrial ribosomes" of Fig. 5C were prepared from the sediment obtained by centrifugation of the postnuclear supernatant of Fig. 5A. It is observed that the absorbance profiles of the ribosomes prepared from the postnuclear and postmitochondrial supernatants (Fig. 5, A and B) are essentially the same; the radioactivity profiles are also similar with the major peak of radioactivity in the disome region. While the total amount of radioactivity incorporated into the disome region is similar for the postnuclear and postmitochondrial supernatants, the specific activity, i.e. counts per min of 14C-arginine per absorbance at 260 nm, in the disome region is higher in the postmitochondrial ribosomes than in the postnuclear ribosomes.
This suggests that additional mitochondrial ribosomes deliberately released by detergent treatment of the postnuclear supernatant are not contributing to the synthesis of protamine; otherwise, the specific aciivity as well as the total incorporation into the disome region of the postnuclear supernatant ribosomes would be higher than with the postmitochondrial supernatant ribosomes. Further, it is seen that the ribosomes prepared directly from a crude mitochondrial fraction incorporate little radioactivity with low specific activity in the disome region (Fig. 5C) of testis ribosomes during testis development was approached by examining the distribution of 14C-arginine and 14C-lysine incorporated into nascent proteins on polysomes prepared from whole testis cells pulsed for 30 min with the two amino acids. In Fig. 6, I to IV, it may be seen that incorporation of each basic amino acid follows a definite pattern from September to December which varies at each stage of development.
In particular, with October and November testis (Fig. 6, 11 and III), the disome fraction, sedimenting at 120 S, incorporates arginine but very little lysine. This observation is consistent with the previous location of protamine synthesis in this fraction (I) since two-thirds of the total residues of protamine are arginines and its lysine content is negligible. 2 This lack of lysine label in the disome region sug-  gests that testis disomes are not making any other basic protein and may be exclusively engaged in protamine synthesis. The incorporation of label into the larger polysome areas also changes at different stages of development.
In September (Fig. 6, 1) there is little incorporation into the larger polysomes but by November (Fig. 6, III)  been characterized but it is not likely to be associated with nascent histones since there is very little histone synthesis at the protamine stage of development (I).
In December (Fig. 6, IV), when the testis consists largely of late spermatids and mature spermatozoa, few polysomes remain and protein synthetic activity has almost ceased.
Effect of Actinomycin D on Protamine Synthesis-Actinomycin D, at low concentrations, specifically inhibits the synthesis of RNA in animal cells (24). Preliminary observations (25) indicated that the incorporation of 14C-arginine into protamine was not inhibited by actinomycin D at a time when protamine synthesis was rapid (cf. protamine stage, Table I). These results were interpreted to suggest that protamine is synthesized on a stable mRNA template formed at an earlier stage of development. In a reexamination of these findings, testis cell suspensions were prepared from trout, at a stage of spermatogenesis at which protamine synthesis had just begun (early protamine stage, Table I) and the effect of actinomycin D on the incorporation of W-arginine into protamine was determined (Table II). At this stage of maturation (September) actinomycin D (0.01 mM) inhibits incorporation by 23 ye, suggesting that the synthesis of some protamine mRNA is still taking place although at later times (October and November) there was no inhibition, indicating that mRNA synthesis had ceased. In September, the number of disomes (Fig. 1) is small but increasing rapidly to the peak observed in October and November when protamine synthesis is maximal.
A reasonable interpretation of this data is that protamine mRNA synthesis occurs just before the onset of protamine synthesis in September and that, once formed, it binds ribosomes to form a stable disome complex on which protamine synthesis can take place.
Issue of June 25, 1970 V. Ling and G. H. Dixon DISCUSSION The data presented here support the conclusion of a previous communication (1) that disomes in the cytoplasm of rainbow trout testis cells are involved in the synthesis of protamine. Characterization of these disomes indicates that they are neither the result of Mg++-induced aggregation of monosomes (Fig. 2) nor the breakdown products of large polysomes (Fig. 4). The testis disomes are different from larger polysomes in that they are resistant to a mild treatment with pancreatic ribonuclease (Fig. 3).
While it is generally assumed that the breakdown of polysomes to monosomes following a mild treatment with ribonuclease is the result of enzymatic degradation of messenger RNA connecting ribosomes in the polysome complex (16,18,26), the resistance of testis disomes to ribonuclease does not preclude the existence of a messenger RNA in the disome but could be due to one (or more) special circumstances.
First, in order to accommodate two ribosomes on the very small putative mRNA for protamine (100 nucleotides (I)), tighter than normal packing might be required and the messenger RNA could be effectively shielded from ribonuclease attack.
In support of this hypothesis it has been shown, for example, that ribosomes bound to synthetic mRNA (poly U) or viral messenger (f2) are able to protect certain portions of the polynucleotide from degradation by pancreatic ribonuclease (27,28).
Second, since protamine is so arginine-rich, its messenger RNA must contain predominantly codons for arginine. Two sets of codons have been found to code for arginine, one completely degenerate, CGX, where X can be any of the four bases, and the other partially degenerate, AGY, where Y is either A or G. Although there is no direct evidence of which set is present in protamine messenger RNA, one piece of indirect evidence suggests that the CGX set may be more probable.
Salmonid and the closely related clupeiform fishes (29) produce protamines which are extremely arginine-rich and contain either no lysine or at most minute traces of it .2 The codons for lysine are AA4 or AAG, which are related by single transition (purine + purine) base changes to the AGA and AGG codons for arginine.
If AGA or AGG were to code for arginine in protamine, then it would seem likely that single step mutations (30) would have occurred during evolution to give rise to mutant lysine-containing protamines.
That lysine does not occur in Salmonid and clupeiform protamines suggests that, in fact, the other codon set, CGX, may be predominant and, in this case, mutation of arginine to lysine would require at least two steps and would, therefore, be much less probable.
It is also interesting to note that codons for several of the limited range of neutral amino acids in protamine, namely proline (CCX), serine (AGcU), and glycine (GGX), are also related by single base changes to the CGX codons for arginine.
When the sequence of nucleotides for the putative mRNA for protamine is written with CGX as the arginine codon, the resulting mRNA is extremely CG-rich and its sequence allows a high degree of secondary structure if written in a "hairpin" structure. Such an mRNA, particularly when in complex with ribosomes in the disome complex, might well be highly resistant to ribonuclease.
The stability of the disomes and the general lack of RNA (29, 31) during the later stages of spermatogenesis raise the question of when the messenger RNA for protamine is synthesized.
The appearance of a large number of disomes in trout testis simultaneously with rapid protamine synthesis (Fig. 1, B and C, and Table I) suggests that protamine messenger RNA and ribosomes combine immediately to form active disomes and that the ratelimiting step for the onset of protamine synthesis is the appearance of pro&mine messenger and formation of active disomes. The resistance of disomes both to low Mg++-induced degradation (Fig. 2) and mild ribonuclease treatment (Fig. 3) suggests that, once synthesized and in complex in disomes, protamine mRNA is very stable. This stability is also reflected by the insensitivity of protamine synthesis to inhibition by actinomycin D once the disome population has become appreciable as in October and November testis (Fig. 1).
A striking feature of spermatogenesis is the marked reduction in the amount of cytoplasm in the differentiating spermatid cell first noted by Caspersson (9), who measured the progressive disappearance of ribonucleic acid from the developing sperm cell by cytochemical and spectrophotometric means. Daoust and Clermont (10) have also shown that, in the rat testis, the RNA of the spermatid is collected into larger and larger granules until they are set free as "residual bodies" shortly before the mature spermatozoa are released into the lumen of the seminiferous tubules.
Lacy (II), studying these residual bodies in greater detail at various stages of their formation, has noted that, in addition to a mass containing numerous RNA particles, these bodies include mitochondria, Golgi membrane, and vesicles, and, also, a concentration of cytoplasmic membranes. Clearly, therefore, a mechanism exists for the removal of cytoplasm from the maturing sperm cell. Consistent with these observations, the data summarized in Table I also indicate that there is a progressive loss of ribosomes from the cytoplasm of developing rainbow trout sperm cells. It may also be seen from the table that the total number of ribosomes present in the testis does not reflect the rate of protamine synthesis in that during October and November the rate of protamine synthesis increased while the ribosome content decreased. Further, in Fig. 1 it may be seen from the changing profile of testis polysomes during development that the proportion of disomes active in the synthesis of protamine increases at the stage of rapid protamine synthesis.
This suggests that the decrease in numbers of ribosomes in the testis cells is not strict1.y the result of a general decrease in the cell cytoplasm but may be the result of specific degradation of certain ribosomes or polysomes not required for further protein synthesis after a particular stage of testis maturation. In this respect, the removal of cytoplasm from the developing sperm cell is probably a specific and well regulated process. In accord with this idea of selectivity in removal of cytoplasm, it has been noted in the developing salmon testis (Oncorhyncus nerka), for example, that the decrease in total RNA was accompanied by a change in its base composition, the mature testis RNA being richer in guanosine (12,31).
It is generally thought that specific functional polysomes are formed for the synthesis of specific proteins at different stages of cell development.
In synchronized HeLa cells, for example, distinct peaks of small cytoplasmic polysomes are observed when ribosomes obtained during the S phase of the cell cycle are fractionated on a sucrose gradient (32). These polysomes are not present at other stages of cell development and have been shown to be associated specifically with the synthesis of histones which is, in turn, coupled with DNA synthesis.
In the trout testis also, striking changes in the polysome profiles are observed (Figs. 1 and 6, I to IV) and reflect the functional state of the tissue in that large numbers of disomes are present in testis cells at a stage of development when the synthesis of protamine is at its peak.