Fur. J . Biochern. SY, 43-50 (1975)
Chemical Synthesis of Partially and Fully Phosphorylated Protamines Lothar WILLMITZER and Karl G . WAGNER Abteilung Molekularbiologie, Gesellschaft fur Molekularbiologische Forschung, Stiickheirn-Braunschweig (Received June 27/August 4, 1975)
Chromatographically purified components Z and YI of clupeine from herring have been phosphorylated by a chemical method. To enhance solubility in trimethylphosphate, the protamine capronate salts were used and phosphorylation was performed with POCI,. Both serine and threonine residues were phosphorylated ; however, besides monophosphate esters considerable amounts of polyphosphate esters were obtained. Its nature was identified by 31P nuclear magnetic resonance spectroscopy and by chemical methods, such as comparison of the hydrolysis rate with that of ADP and phosphocreatine, respectively. The pyrophosphate bonds were cleaved by controlled hydrolysis with HCI. The phosphorylated protamine species were purified by gel filtration and chromatography on a Sephadex CM column. Fully phosphorylated clupeine Z (3 serine residues and 3 phosphates) and YI (3 serine, 2 threonine residues and 5 phosphates) were obtained as well as partially phosphorylated fractions; in the case of clupeine Z the resolution into mono, doubly and triply (fully) phosphorylated fractions was excellent. The overall yield of phosphorylated clupeine was better than 50 7:. The mature sperm cell nuclei of certain families of fish contain relatively small basic proteins, the protamines, which are unique in amino acid composition and also show some common features of residue sequences [ 1,2]. More than two-thirds of the residues are arginine mainly arranged in clusters and the remainder consists of aliphatic, inert residues like proline, glycine, alanine or valine and the aliphatic, hydroxylic residues serine and threonine. During spermatid maturation the histones are replaced by the protamines. In this process the protamines are intensively phosphorylated and dephosphorylated [3]. controlled phosphorylation has been supposed to be involved in the correct binding of protamines to DNA and the displacement of the histones, whereas dephosphorylation has been related to condensation of spermatid chromatin [4]. If reasonable amounts of phosphorylated protamine species were available the investigation of these processes in vitro would be greatly facilitated. The influence of serine or threonine-bonded phosphates upon the binding of protamines to DNA and chromatin could be studied, as well as its effect upon removal of histones. Furthermore phosphorylated protamines are -~
Part of this work was presented at the Friihjahrstagung der Gesellschafifur Biologischr Chemie in Heidelberg, 1975 and at the V th International Biophysics Congress in Copenhagen, 1975. Abbreviation. NMR. nuclear magnetic resonance. Eqyme. Alkaline phosphatase (EC 3.1.3.1).
suitable substrates for the nuclear protein phosphatases [5] and removal of phosphate from complexes formed with DNA or chromatin could be investigated. Recently, the preparation of phosphorylated protamines has been described by extraction from hormonally induced trout testes at the early stage of development. As protamines consist of several species with different amino acid composition and different serine and threonine content. Phosphorylation in vivo leads to complex mixtures of species differing in phosphate content and amino acid composition. The resolution of such mixtures is very difficult [3]. Phosphorylation in vitro definitely improves this situation as one can start with homogeneous (unphosphorylated) protamine species purified by ionexchange chromatography [6].Using chemical methods for phosphorylation the possible preparation of large quantities of phosphorylated protamines is a further advantage, moreover the desired degree of phosphorylation can be easily controlled. It is shown in the present work that the resolution into species, only differing in the number of phosphorylated serine or threonine residues, can be performed. EXPERIMENTAL PROCEDURE Materials
Capronic acid was a product of Fluka Feinchemikalien GmbH, Neu-Ulm ; trimethylphosphate and
44
Chemical Phosphorylation of Protamines
Table 1. Amino acid composition of clupeine Z and of phosphorylated clupeine Z The data obtained were related to alanine. Phosphorylated clupeine Z was taken from fraction I of Fig. 4 Amino acid
Residues per clupeine molecule -~
expected
- _ _ ~ -____
found for __
- -
clupeine Z Alanine Serine Proline Valine Arginine
3.0 3.0 2.0 2.0 21
3.0 2.1,
2.1, 2.0* 22.3
__
--
__ __
phosphorylated clupeine Z 3.0 2.5, 1.9, 2.0, 21.3
phosphoric oxychloride were purchased from Merck Darmstadt and were distilled before use. O-PhosphoL-serine and 0-phospho-DL-threonine were obtained from Sigma Chemical Co. and N-phosphocreatine from Merck Darmstadt. Clupeine sulfate was purchased from Carl Roth OHG, Karlsruhe; pure clupeine fractions were obtained by chromatography on Sephadex C25 according to Ando and Watanabe [6]. The results of the amino acid analysis are shown in Table 1. Alkaline phosphatase from calf mucosa with a specific activity (p-nitrophenyl phosphate) of 350 units/mg protein was purchased from Boehringer Mannheim. Phosphate Assays
Total phosphate was determined by the method of Ames [7] ; acid-labile phosphate' was assayed after heating the sample for 20 min in 1 N HC1 at 100 "C, base-labile phosphate after treatment with 0.5 N NaOH at 44 "C for 16- 18 h [8]. Inorganic phosphate liberated by these treatments was routinely determined according to Martin and Doty [9] using 3 M trichloroacetic acid instead of silicotungsticacid and subsequent centrifugation of the precipitated protein. Control experiments indicated, however, that because of protein precipitation only 80 to 90 % of the inorganic phosphate present could be detected. Hence for precise determinations, a fast method for deproteination was elaborated: samples of about 1.5 ml were applied onto a short (1.5 x 5.5 cm) Dowex 50 WX 8 (H+-form, 100- 200 mesh) column and washed through with 0.05 N HCl. Protein was completely absorbed, whereas inorganic phosphate was obtained without any loss.
suspended in 50 ml freshly distilled trimethylphosphate by gentle stirring. To the resulting fine suspension 100 p1 POCl, (about 1 mmol) were added, thereafter the solution became completely clear. Depending on the degree of phosphorylation desired, stirring was continued for 2 to 6 days at room temperature. Usually the solution became turbid after 2 to 3 days by precipitation of clupeine, then another 30- 50 pl POCl, were added to clear the solution. The reaction was terminated by pouring the mixture into twice the volume of water. Trimethylphosphate was removed by adsorbing the clupeine onto a SephadexC25 column (2.5 x 6.5 cm), washing with water until the ultraviolet absorption (224 nm) of the eluate was negligible and the clupeine was detached with 2 M NaC1. Phosphorylated clupeine was desalted on Biogel P2, lyophilized and then dissolved (20 mg/ml) in 6 N HCl for 6 to 8 h at room temperature in order to split off the acidlabile phosphates. After neutralization with 3 N NaOH under vigorous stirring and ice-cooling, intact clupeine was separated from fragments obtained (acid treatment) by gel filtration on Biogel P4 (200- 400 mesh). A column of 2.5 x 50 cm was used for about 50 mg phosphorylated clupeine. Elution was performed with 0.5 M acetic acid with a flow rate of 20 ml/h. The high-molecular-weight fractions were collected and lyophilized. Fractionation according to degree of phosphorylation was performed by Chromatography on a Sephadex C25 column (1.6 x 16 cm) and stepwise elution as indicated in Fig. 5. Protein Assays
Protein concentration was determined by the biuret method [lo]. Free amino acids were detected by fluorescamine [l 11; amino acid composition was assayed on a Biocal BC200 amino acid analyser. The relative content of intact protamines and cleavage products produced by HC1 treatment was obtained by determining the relative content of N-terminal amino acids [12]. N-terminal group determination was performed by dansylation according to Grosse and Labouesse [ 131; dansylamino acids were separated by two-dimensional thin-layer chromatography (cf. [141, solvent A and B). The fluorescent spots of the dansylamino acids were removed from the plates, eluted by a mixture of chloroform, methanol and acetic acid and the fluorescence was measured [13].
Phosphorylation of' Clupeine
Spectroscopic Methods
To enhance solubility clupeine chloride obtained after fractionation was converted to clupeine capronate by dissolving in 0.25 M sodium capronate and gel filtration on Sephadex G25. For phosphorylation 100 mg clupeine capronate (about 20 pmol) were
31Pnuclear magnetic resonance (NMR) spectra were recorded on a Varion XL-100 spectrometer with 850/0 phosphoric acid as external standard; all solutions contained 0.2 mM EDTA. Fluorescence measurements were done on a Schoeffel RRS fluorometer.
45
L. Willmitzer and K. G. Wagner Table 2. Total. base-labile and acid-labile phosphate conlent of’ clupeine Z phosphorylated for dq’erent reacrions times For experimental conditions cf. Experimental Procedure Time
Phosphates in clupeine molecule ~-
total
-
base-labile
h
acid-labile
total ..~
36 60 120
1.5 2 5
42 36 10
16 35 52
be formed during base treatment would fail to form a coloured adduct in the molybdenum blue reaction [18]. Evidence that inorganic pyrophosphate is formed could be obtained from a comparison of the rate of molybdenum blue formation in the presence of cysteine [18]. In principle, the formation of phosphorylated arginyl residues could be considered as well. However, under the reaction conditions applied the guanidinium group should be protonated and hence be inaccessible to an electrophilic attack. Nature of the Acid-Labile Phosphate
RESULTS AND DISCUSSION
Rate of Phosphorylation Under the conditions applied and described in Experimental Procedure phosphorylation proceeds rather slowly; the reasons may be the rather low concentrations used and/or low reactivities due to steric or electronic properties of the serine and threonine residues. It is rather peculiar to this reaction that the amount of phosphate covalently bound was found to be a linear function of time, even for long reaction times and up to high degrees of phosphorylation. About 6 mol phosphate per mol of clupeine were incorporated after six days. Since unfractionated clupeine contains on the average 3 to 4 hydroxyl groups per molecule clupeine, one has to elucidate the nature of the excess 2 to 3 phosphates. Acid-Labile and Base-Labile Phosphate The phosphate esters of serine and threonine are known to be acid-stable and base-labile [15]. Table 2 shows data for the different amounts of base-labile and acid-labile phosphate obtained by phosphorylation of unfractionated clupeine. The relative amount of acid-labile phosphate increases with reaction time. It is obvious that in any case acid-labile and baselabile phosphate do not add up to 100 %. Breyer et al. [161 reported phosphorylation of poly(DL-serine) by chlorophosphoric acid and also obtained a considerable amount of acid-labile phosphate. They proposed that acid-labile phosphatases were linked to the a-amino groups (N-a-phosphates), free a-amino groups were obtained after N-0-acyl migrations which involve the peptide carboxyl group and the serine or threonine side-chain OH [17]. A different explanation is the formation of serine and threonine polyphosphate esters; this is supported by the linear rate of phosphorylation and by the time dependency of the formation of acid-labile phosphates and especially by the fact that considerable amounts of phosphate are neither identified as base-labile nor acid-labile (Table 2). Polyphosphates which would
The nature of the acid-labile phosphate was investigated by observing the time course of phosphate liberation in 1 N HC1 at 44 ”C by comparison with different phosphoryl compounds. The results are shown in Fig. 1 which also contains the data reported for phosphorylated poly(m-serine) [16]. N-Phosphocreatine is hydrolysed very quickly indicating that phosphoarginyl bonds should not exist in the present phosphoclupeine; N-a-phosphates, when present, should be split even faster than phosphocreatine. The time courses of phosphate liberation from 5’-ADP, used as a model for polyphosphate esters, phosphorylated clupeine and phosphorylated poly(DL-serine) are very similar, indicating that the acid-labile phosphate is bound similarly in these compounds. The occurrence of the N-0-acyl migration, a prerequisite for the existence of acid-labile N-a-phosphates, can be checked. The ester bonds formed hereby can be split by treatment with 6 N HCl giving rise to peptide fragments with serine as the N-terminal amino acid. A reaction time of 4 h at room temperature was reported to be sufficient to split 50% of possible ester bonds present in clupeine [ 191. Phosphorylated clupeine Z, containing alanine as N-terminal amino acid was incubated for 4 h in 6 N HCl at room temperature and the resulting N-terminal residues were determined by dansylation. Only about one N-terminal serine per 15-N-terminal alanines was found. Assuming the acid-labile phosphate present in the phosphorylated clupeine to be N-a-phosphates, two ester bonds would have been expected (about 2 acid-labile phosphates per molecule were present in the clupeine) thus giving rise to equal amounts of N-terminal serine and alanine after 50% hydrolysis. The rather low value found for serine makes the existence of N-a-phosphates very unlikely. 31
P-NMR Spectroscopy
The nature of the phosphate bond present in phosphoryldted clupeine was elucidated by 31P-NMR spectroscopy. Phosphorylated unfractionated clupeine was used, in order to supply enough material. The
46
Chemical Phosphorylation of Protamines
I
0
I
5
20
15 10 Hydrolysis time ( h )
25
-
Fig. 1. Rate of cleavage of acid-labilephosphatefromphosphorylated clupeine and several related substances. Batches containing total phosphate of 0.7 mM (0,ADP), 0.1 mM (0,N-phosphocreatine) and 0.3 mM (A, phosphorylated unfractionated clupeine containing about 2.5 phosphates per molecule clupeine) were incubated in 1 N HCl at 44 "C and aliquots of 1 ml were withdrawn at different times for determination of inorganic phosphate as described in Experimental Procedure. The data for phosphorylated poly(DL-serine) (0)were taken from the literature [16]. All the data were related to the total amount of acid-labile phosphate obtained after an incubation time of 24 h
++ + 0 I
0 I
0 I
I o=p-o-
j2+3
4 pn2.5 I
I
I
0
I
0Formulae of ( a ) serine monophosphate residue; ( 6 ) serine dbhosphate residue and ( c ) serine oligophosphate residue. The phosphoryl groups exerting equal (or practically equal) chemical shifts in 31PNMR spectroscopy are designated by the same number
assignment of the four signals detected was performed through the pH dependency of their chemical shifts as is schematically shown in Fig. 2A. Two of the signals are shifted to lower field when the pH is raised, whereas the other two are scarcely affected. Hence the former signals must arise from phosphates with two ionizable -OH groups, the latter from phosphates possessing only one -OH group. Fig. 2B indicates the chemical shifts and their pH-dependence of reference compounds such as serine phosphate, ADP and ATP (chemical shifts of ADP and ATP are taken from [20]). A comparison of Fig. 2A and 2 B clearly shows that location and pH-dependence of signal 1 (Fig. 2A) closely resembles that of serine phosphate (Fig. 2B), signal 2 resembles those of P (/?)of ADP and P (y) of ATP, signal 3 those of the P (a) of ADP and ATP, respectively, and signal 4 that of P (fi) of ATP. This strongly supports the conclusion that phosphorylated clupeine contains monophosphates of serine (and threonine) and polyphosphate esters
pH 7.2
I 0
I
+ 10
I +x)
I 30
Chemical shift (ppm)
Fig. 2. Schematic positioning of the I' P resonances ofphosphorylated clupeine and of phosphoserine, N-phosphocreatine, A D P and ATP at different p H values. Concentration of clupeine was 20 mM containing about 3 phosphates per clupeine molecule. Clupeine was dissolved in 6 M urea containing 0.2 mM EDTA to enhance solubility. The number of scans was 9000, the pH was adjusted with HCI or NaOH. (A) "P resonances of phosphorylated clupeine, the numbers refer to the phosphoryl groups illustrated in the formulae, (B) "P resonances of the reference compounds; phosphoserine (-.-.-)and N-phosphocreatine (....-.) were recorded with a 0.1 M solution in 0.2 mM EDTA. the number of scans were about 10 to 20; the spectra of ADP (----) and ATP (-) were taken from the literature [20]; these had been recorded at pH values of 2.2, 6.15 and 6.9 in the case of ADP and of 3.6, 6.15 and 7.15 in the case of ATP. The chemical shifts are related to external 85 phosphoric acid
47
L. Willmitzer and K. G. Wagner
w
I
,
1
I
10
x)
30
40
Fraction number
Fig. 3. Evidence that in clupeine both serine and threonine residues ure phosphoryhted. 40 mg of phosphorylated unfractionated clupeine containing about 2 mol of phosphate per mol of protamine were hydrolysed with 2 N HCI for 8 h at 104 "C (total volume 3 ml), applied onto a small Dowex 50 WX 8 column (H+ form, 1.6 x 28 cm), and eluted with 0.05 N HCI [21]. The fractions (4 ml) were checked for inorganic phosphate (A), organic phosphate (0)and a-amino groups with fluorescamine (0)
of these hydroxy amino acids, the latter constitute the acid-labile phosphates. This conclusion is strengthened by the findings that treatment of phosphorylated clupeine with 1 N HC1 at 44 "Cfor 14 h (cf. Fig. 1) and removal of inorganic phosphate by gel filtration leaves phosphorylated clupeine with a 31P-NMR spectrum, which contains only signal 1 (serine and threonine phosphate) whereas the signals 2 to 4 have vanished. The Phosphorylated Amino Acid Residues
In order to clarify whether, under the reaction conditions employed, both serine and threonine residues are phosphorylated, unfractionated clupeine was phosphorylated, hydrolysed by 2 N HCl (104 "C, 8 h) and lyophyilized. By the method proposed by Schaffer et al. [21] the hydrolysate was applied onto a small Dowex 50 (H' form) column, after elution of the acid residues and phosphate with 0.05 N HCI, the fractions were tested for phosphate and for aamino groups with fluorescamine. The results shown in Fig. 3 indicate a preceding inorganic and two organic phosphate peaks, the latter are fluorescamine positive. Running the column with authenic substances the first fluorescamine positive peak was assigned to phosphoserine and the second to phosphothreonine. These results, further supported by high-voltage paper electrophoresis, show that under the present experimental conditions both serine and threonine residues are phosphorylated.
Cleavage of Pyrophosphate Bonds
Preparation of phosphorylated protamines, containing only monophosphate esters of serine or threonine, requires cleavage of the pyrophosphate bonds present. This was done by treatment with HCI; however, to reduce concomitant cleavage of peptide bonds, the conditions of Fig. 1 were not applied but hydrolysis was performed at room temperature with 6 N HC1. Following the time course of phosphate liberation it was found that the reaction is completed after about 6 h. The occurrence of peptide bond cleavage is manifested by the appearance of new N-terminal amino acids. This was checked with clupeine Z which has alanine as the N-terminal residue, two internal alanines, 3 serines and 21 arginines. Treatment of phosphorylated clupeine Z for 6 h with 6 N HCl and subsequent N-terminal amino acid analysis revealed 1 - 2 serine residues and 2- 3 arginine residues per 15 alanine residues. These results indicate that during acid treatment partial peptide bond cleavage occurred. Chromatographic Fractionation of Phosphorylated Clupeine
Intact phosphorylated clupeine Z was separated from peptide fragments by chromatography on Biogel P4 as shown in Fig. 4. Analysis of the N-terminal amino acids (cf. the legend of Fig. 4) reveals that the first peak emerging consists of about 95 "/, intact clupeine Z. The amount of N-terminal alanine can be
Chemical Phosphorylation of Protarnines
48
-
20
10
30
Fraction number
Fig. 4. Gelfiltrotion ofphosphorylufed clupeine 2 after cleuwge of'rhe p,vropphosphure bonds. As describcd in Experimental Procedure clupeine fraction Z (50 mg), after phosphorylation and cleavage of the pyrophosphate bonds, was fractionated on Biogel P4,in order to separate peptide fragments obtained during acid treatment. The fractions (10 ml) were collected and combined as indicated in the figure and the N-terminal amino acids were determined as described in Experimental Procedure. The following numbers of N-terminal serine and arginine related to 50 N-terminal alanine residues (alanine is the N-terminal residue of intact clupeine Z) were found: fraction I 2 arginine and less than 1 serine; fraction I1 5 argininc and 2 serine; fraction 111 23 arginine and fraction IV 41 arginine and 6 serine residues
0.10
-
. E
0
-._5E 0.05 -% 0
10
20 Fraction number
Fig. 5. Sepurution of phosphorylufed clupeine 2 according to the dtgree qf'phosphor~,lution. Fraction 1 from Fig. 4 (about 30 mg protein) was lyophilized, dissolved in about 10 ml of 0.05 M sodium acetate of pH 5.7 containing 1.2 M NaCI, applied to a Sephadex C25 column (1.6 x 16 cm) and the column was washed with the same buffer. Phosphorylated clupeine Z components were eluted (20 ml/h) by the following stepwise NaCl gradients (containing 0.05 M sodium acetate of pH 5.7): fraction 8 to 18 1.35 M ; fraction 18 to the end 1.5 M NaCI. Fractions of 10 ml were collected and andlysed for organic phosphate (A) and protein content (0).The inserted small figure indicates the number of phosphates per clupeine Z molecule determined for the different fractions as described in Experimental Procedure
taken as a measure of intact clupeine Z, for peptide bond cleavage at the alanine residues can be neglected, as there are only two internal alanines per 31 residues. Furthermore these bonds are more stable than serine peptide bonds [22]. Amino acid analysis of fraction I of Fig. 4 as shown in Table 1 also confirms that it consists mostly of intact clupeine. The reduction of the serine value usually found with acid hydrolysis [23] is further augmented due to the presence of 0phosphoserine residues [24].
Fraction I of Fig. 4 and the respective fractions obtained by phosphorylating clupeine YI were reanalyzed to determine the nature of the phosphate bond. The amount of acid-labile phosphate found did not exceed 5 to 6% of the total phosphate; this should be reduced upon slightly prolonging the acid treatment. Separation of the phosphorylated clupeine Z (fraction I of Fig. 4) according to the degree of phosphorylation was performed by chromatography on Sephadex CM using a stepwise gradient of NaCl as
49
L. Willrnitzer and K. G. Wagner
shown in Fig. 5. The appearance of three distinct phosphate-containing fractions with clupeine Z, which has three serine and no threonine residue, suggests that fractions of singly, doubly and triply phosphorylated clupeine were obtained. This is confirmed by determining the phosphate/clupeine ratio; the results are indicated at the top of Fig. 5. The triply phosphorylated clupeine should accordingly represent a homogeneous fraction of phosphorylated clupeine, while the doubly and singly phosphorylated fractions may represent mixtures of the respective isomers. The phosphorylation of clupeine YI, which contains 3 serine and 2 threonine residues, was performed in a similar way. After chromatographic resolution according to the degree of phosphorylation a minor fraction containing 5 phosphates per molecule and fractions with less phosphate were obtained.
Enzymic Cleavage of Phosphate Bonds To demonstrate that the phosphate esters of phosphorylated clupeine can be cleaved enzymically, phosphorylated fraction YI was treated with alkaline phosphatase from calf mucosa. 1 mg clupeine Y1 containing about 4 phosphates per molecule was incubated with 20 pg phosphatase for 30 min at 30 "C in 1 ml of 5 0 m M Tris, 1 mM dithiothreitol, 0.1 M NaCl and 2 mM MgC1, of pH 7.5. After removal of the proteins as described in Experimental Procedure 2506 of the input phosphate was determined as inorganic phosphate. This findings are consistent with those of Meisler and Langan [ 5 ] , who reported that this alkaline phosphatase releases phosphate from enzymically phosphorylated salmine at a strongly reduced rate relative to low-molecular-weight substrates such as glycerol 2-phosphate.
CONCLUSIONS The present work describes the first investigation
to obtain homogeneous fractions of phosphorylated protamines by a chemical method. After completion of this work Ullman and Perlman published a paper [25] on chemically phosphorylated protamines, which they used as substrates for phosphoprotein phosphalases. However, they started from unfractionated protamine and succeeded in incorporating only about one mole of phosphate per mole of protamine by reaction with inorganic phosphate and trichloroacetonitrile. They did not try to fractionate the mixture of different phosphorylated species obtained. Enzymic phosphorylation of protamine was described by Meisler and Langan [ 5 ] ; starting from unfractionated protamine they also reported about one mol of phosphate incorporated into one mol of protamine by a histone kinase. Phosphorylation in vivo by
hormonal induction of trout testes and extraction of phosphorylated protamines has been mentioned above. Enzymatic phosphorylation certainly circumvents the formation of polyphosphate esters obtained by the chemical methods. However, it is easy to split the pyrophosphate bonds. Thus, the chemical method, as described in the present work, has the following important advantages : possibility of phosphorylating large amounts of protamines, control of the degree of phosphorylation and possibility of obtaining highly and practically fully phosphorylated species. By analytical procedures and by ,lP-NMR spectroscopy the nature of the acid-labile phosphate obtained by chemical phosphorylation was elucidated and found to originate from polyphosphate esters of serine and threonine. A comparison of the rate of cleavage of the acid-labile phosphate obtained by phosphorylation of poly(DL-serine) with chlorophosphoric acid in the work of Breyer et al. [16] with that of phosphorylated protamines of the present work strongly favors the suggestion that phosphorylated poly(m-serine) contained serine polyphosphate esters and not N-a-phosphates; the latter should be hydrolysed very quickly in acid conditions, faster than creatine phosphate (cf. Fig. 1). It is also likely that in the phosphorylated protamine of the work of Ullman and Perlman [25] the acid-labile phosphate found is serine polyphosphate and not phosphoarginine as the authors suggest. The phosphorylation with POCl, in anhydrous trimethylphosphate constitutes a mild procedure for phosphate incorporation into protamine as is demonstrated by the slow rate of this reaction. An essential prerequisite for the success of this procedure was the fact that the hydrophobic counterion capronate was used. Together with the action of the applied POCl, this leads to complete dissolution of protamine. Application of chloride or acetate salts of protamine resulted in turbid suspensions and a failure to obtain markable phosphate incorporation. The circular dichroic spectrum recorded with protamine capronate in trimethylphosphate after addition of POCl, revealed that about 502; of the peptide bonds are in a-helical conformation, a situation which is only obtained by dissolving protamine in a strong a-helix-forming solvent like dichloroethanol [26]. We are very gratcful to D r M.-R. Kula for performing the iimino acid analysis and to Dr V. Wray for measuring thc magnetic rcsonance spectra and for linguistic advice. This work was supported by the Bunrlesniiiilsterium fur Forscllung unri 7 i d i n d o g k (RCT 16) and the Dwische F~~rschun~ge~ntc~in.s~liuf/ (Wa 91 15. 7).
REFERENCES 1. Ando, T., Yamasdki. M. & Suzuki, K. (1973) Prolamine.s,
Springcr-Verlag, Berlin-Heidelberg-New York. 2. Bretzel. G. (1972) Hoppe-Seyler's Z. Physioi. ('hem. 353, 933 943.
50
L. Willmitzer and K. G. Wagner: Chemical Phosphorylation of Protamines
3. Louie, A. J. & Dixon, G. H. (1974) Can. J. Biochem. 52, 536546. 4. Louie, A. J. & Dixon, G. H. (1972) J. Biol. Chem. 247, 79627968. 5 . Meisler, M. H. & Langan, T. A. (1969) J. Biol. Chem. 244, 4961 -4968. 6. Ando, T. & Watanabe, S. (1969) Int. J. Protein Res. I , 221 - 224. 7. Ames, B. N. (1966) Merhods Enzymol. 8, 115-118. 8. Ahmed, K. & Judah, J. D. (1967) Methods Enzymol. 10, 777780. 9. Martin, J. B. & Doty, P. M. (1949) Anal. Chem. 21, 965-967. 10. Zamenhof, S. (1957) Methodr Enzymol. 3, 702. 11. Bernardo, S., Weigele, M., Toome, V., Manhart, K., h i m gruber, W., Bohlen, P., Stein, S. & Udenfriend, S. (1974) Arch. Biochem. Biophys. 163, 390- 399. 12. Spivak, V. A., Levjant, M. I., Katrukha, S. P. & Varshavsky, J. M. (1971) Anal. Biochem. 44, 503-518. 13. Gros, C. & Labouesse, B. (1969) Eur. J. Biochem. 7,463-470. 14. Seiler, N. & Wiechmann, J. (1964) Experientia (Busel) 20, 559- 560.
15. Anderson, L. & Kelley, J. J. (1959) J. Am. Chem. Sor. 81, 2275 - 2216. 16. Breyer, U., Lapidot, Y., Kurtz, J., Bohak, Z. & Katchalski, E. (1967) Monursh. Chem. 98, 1386-1394. 17. Iwai, K. & Ando, T. (1967) Methods Enzymol. 11,263-282. 18. Flynn, R. M., Jones, M. E. & Lipmann, F. (1954)J. Biol. Chem. 211,791-796. 19. Iwai, K. (1960) Nippon Kagaku Zusshi, 81, 1302. 20. Cohn, M. & Hughes, T. R. (1960) J . Biol. Chem. 235, 32503253. 21. Schaffer, N. K., May, S. M. & Summerson, W. H. (1954) J. Biol. Chem. 206,201 - 207. 22. Hill, R. L. (1965) Adv. Protein Chem. 20, 37- 107. 23. Needleman, S. B. (1970) Protein Sequence Determination, pp. 93 -95, Springer-Verlag, Berlin-Heidelberg-New York. 24. Allerton, S. E. & Perlmann, G. E. (1965) J. Biol. Chem. 240, 3892- 3898. 25. Ullman, B. & Perlman, R. L. (1975) Biochem. Biophys. Res. Commun. 63,424-432. 26. Suzuki, K. & Ando, T . (1968) J . Biochem. (Tokyo) 63, 403405.
L. Willmitzer and K. G. Wagner, Gesellschaft fur Molekularbiologische Forschung mbH, D-3301 Stockheim iiber Braunschweig, Mascheroder Weg 1, Federal Republic of Germany
Chemical Synthesis of Partially and Fully Phosphorylated Protamines Lothar WILLMITZER and Karl G . WAGNER Abteilung Molekularbiologie, Gesellschaft fur Molekularbiologische Forschung, Stiickheirn-Braunschweig (Received June 27/August 4, 1975)
Chromatographically purified components Z and YI of clupeine from herring have been phosphorylated by a chemical method. To enhance solubility in trimethylphosphate, the protamine capronate salts were used and phosphorylation was performed with POCI,. Both serine and threonine residues were phosphorylated ; however, besides monophosphate esters considerable amounts of polyphosphate esters were obtained. Its nature was identified by 31P nuclear magnetic resonance spectroscopy and by chemical methods, such as comparison of the hydrolysis rate with that of ADP and phosphocreatine, respectively. The pyrophosphate bonds were cleaved by controlled hydrolysis with HCI. The phosphorylated protamine species were purified by gel filtration and chromatography on a Sephadex CM column. Fully phosphorylated clupeine Z (3 serine residues and 3 phosphates) and YI (3 serine, 2 threonine residues and 5 phosphates) were obtained as well as partially phosphorylated fractions; in the case of clupeine Z the resolution into mono, doubly and triply (fully) phosphorylated fractions was excellent. The overall yield of phosphorylated clupeine was better than 50 7:. The mature sperm cell nuclei of certain families of fish contain relatively small basic proteins, the protamines, which are unique in amino acid composition and also show some common features of residue sequences [ 1,2]. More than two-thirds of the residues are arginine mainly arranged in clusters and the remainder consists of aliphatic, inert residues like proline, glycine, alanine or valine and the aliphatic, hydroxylic residues serine and threonine. During spermatid maturation the histones are replaced by the protamines. In this process the protamines are intensively phosphorylated and dephosphorylated [3]. controlled phosphorylation has been supposed to be involved in the correct binding of protamines to DNA and the displacement of the histones, whereas dephosphorylation has been related to condensation of spermatid chromatin [4]. If reasonable amounts of phosphorylated protamine species were available the investigation of these processes in vitro would be greatly facilitated. The influence of serine or threonine-bonded phosphates upon the binding of protamines to DNA and chromatin could be studied, as well as its effect upon removal of histones. Furthermore phosphorylated protamines are -~
Part of this work was presented at the Friihjahrstagung der Gesellschafifur Biologischr Chemie in Heidelberg, 1975 and at the V th International Biophysics Congress in Copenhagen, 1975. Abbreviation. NMR. nuclear magnetic resonance. Eqyme. Alkaline phosphatase (EC 3.1.3.1).
suitable substrates for the nuclear protein phosphatases [5] and removal of phosphate from complexes formed with DNA or chromatin could be investigated. Recently, the preparation of phosphorylated protamines has been described by extraction from hormonally induced trout testes at the early stage of development. As protamines consist of several species with different amino acid composition and different serine and threonine content. Phosphorylation in vivo leads to complex mixtures of species differing in phosphate content and amino acid composition. The resolution of such mixtures is very difficult [3]. Phosphorylation in vitro definitely improves this situation as one can start with homogeneous (unphosphorylated) protamine species purified by ionexchange chromatography [6].Using chemical methods for phosphorylation the possible preparation of large quantities of phosphorylated protamines is a further advantage, moreover the desired degree of phosphorylation can be easily controlled. It is shown in the present work that the resolution into species, only differing in the number of phosphorylated serine or threonine residues, can be performed. EXPERIMENTAL PROCEDURE Materials
Capronic acid was a product of Fluka Feinchemikalien GmbH, Neu-Ulm ; trimethylphosphate and
44
Chemical Phosphorylation of Protamines
Table 1. Amino acid composition of clupeine Z and of phosphorylated clupeine Z The data obtained were related to alanine. Phosphorylated clupeine Z was taken from fraction I of Fig. 4 Amino acid
Residues per clupeine molecule -~
expected
- _ _ ~ -____
found for __
- -
clupeine Z Alanine Serine Proline Valine Arginine
3.0 3.0 2.0 2.0 21
3.0 2.1,
2.1, 2.0* 22.3
__
--
__ __
phosphorylated clupeine Z 3.0 2.5, 1.9, 2.0, 21.3
phosphoric oxychloride were purchased from Merck Darmstadt and were distilled before use. O-PhosphoL-serine and 0-phospho-DL-threonine were obtained from Sigma Chemical Co. and N-phosphocreatine from Merck Darmstadt. Clupeine sulfate was purchased from Carl Roth OHG, Karlsruhe; pure clupeine fractions were obtained by chromatography on Sephadex C25 according to Ando and Watanabe [6]. The results of the amino acid analysis are shown in Table 1. Alkaline phosphatase from calf mucosa with a specific activity (p-nitrophenyl phosphate) of 350 units/mg protein was purchased from Boehringer Mannheim. Phosphate Assays
Total phosphate was determined by the method of Ames [7] ; acid-labile phosphate' was assayed after heating the sample for 20 min in 1 N HC1 at 100 "C, base-labile phosphate after treatment with 0.5 N NaOH at 44 "C for 16- 18 h [8]. Inorganic phosphate liberated by these treatments was routinely determined according to Martin and Doty [9] using 3 M trichloroacetic acid instead of silicotungsticacid and subsequent centrifugation of the precipitated protein. Control experiments indicated, however, that because of protein precipitation only 80 to 90 % of the inorganic phosphate present could be detected. Hence for precise determinations, a fast method for deproteination was elaborated: samples of about 1.5 ml were applied onto a short (1.5 x 5.5 cm) Dowex 50 WX 8 (H+-form, 100- 200 mesh) column and washed through with 0.05 N HCl. Protein was completely absorbed, whereas inorganic phosphate was obtained without any loss.
suspended in 50 ml freshly distilled trimethylphosphate by gentle stirring. To the resulting fine suspension 100 p1 POCl, (about 1 mmol) were added, thereafter the solution became completely clear. Depending on the degree of phosphorylation desired, stirring was continued for 2 to 6 days at room temperature. Usually the solution became turbid after 2 to 3 days by precipitation of clupeine, then another 30- 50 pl POCl, were added to clear the solution. The reaction was terminated by pouring the mixture into twice the volume of water. Trimethylphosphate was removed by adsorbing the clupeine onto a SephadexC25 column (2.5 x 6.5 cm), washing with water until the ultraviolet absorption (224 nm) of the eluate was negligible and the clupeine was detached with 2 M NaC1. Phosphorylated clupeine was desalted on Biogel P2, lyophilized and then dissolved (20 mg/ml) in 6 N HCl for 6 to 8 h at room temperature in order to split off the acidlabile phosphates. After neutralization with 3 N NaOH under vigorous stirring and ice-cooling, intact clupeine was separated from fragments obtained (acid treatment) by gel filtration on Biogel P4 (200- 400 mesh). A column of 2.5 x 50 cm was used for about 50 mg phosphorylated clupeine. Elution was performed with 0.5 M acetic acid with a flow rate of 20 ml/h. The high-molecular-weight fractions were collected and lyophilized. Fractionation according to degree of phosphorylation was performed by Chromatography on a Sephadex C25 column (1.6 x 16 cm) and stepwise elution as indicated in Fig. 5. Protein Assays
Protein concentration was determined by the biuret method [lo]. Free amino acids were detected by fluorescamine [l 11; amino acid composition was assayed on a Biocal BC200 amino acid analyser. The relative content of intact protamines and cleavage products produced by HC1 treatment was obtained by determining the relative content of N-terminal amino acids [12]. N-terminal group determination was performed by dansylation according to Grosse and Labouesse [ 131; dansylamino acids were separated by two-dimensional thin-layer chromatography (cf. [141, solvent A and B). The fluorescent spots of the dansylamino acids were removed from the plates, eluted by a mixture of chloroform, methanol and acetic acid and the fluorescence was measured [13].
Phosphorylation of' Clupeine
Spectroscopic Methods
To enhance solubility clupeine chloride obtained after fractionation was converted to clupeine capronate by dissolving in 0.25 M sodium capronate and gel filtration on Sephadex G25. For phosphorylation 100 mg clupeine capronate (about 20 pmol) were
31Pnuclear magnetic resonance (NMR) spectra were recorded on a Varion XL-100 spectrometer with 850/0 phosphoric acid as external standard; all solutions contained 0.2 mM EDTA. Fluorescence measurements were done on a Schoeffel RRS fluorometer.
45
L. Willmitzer and K. G. Wagner Table 2. Total. base-labile and acid-labile phosphate conlent of’ clupeine Z phosphorylated for dq’erent reacrions times For experimental conditions cf. Experimental Procedure Time
Phosphates in clupeine molecule ~-
total
-
base-labile
h
acid-labile
total ..~
36 60 120
1.5 2 5
42 36 10
16 35 52
be formed during base treatment would fail to form a coloured adduct in the molybdenum blue reaction [18]. Evidence that inorganic pyrophosphate is formed could be obtained from a comparison of the rate of molybdenum blue formation in the presence of cysteine [18]. In principle, the formation of phosphorylated arginyl residues could be considered as well. However, under the reaction conditions applied the guanidinium group should be protonated and hence be inaccessible to an electrophilic attack. Nature of the Acid-Labile Phosphate
RESULTS AND DISCUSSION
Rate of Phosphorylation Under the conditions applied and described in Experimental Procedure phosphorylation proceeds rather slowly; the reasons may be the rather low concentrations used and/or low reactivities due to steric or electronic properties of the serine and threonine residues. It is rather peculiar to this reaction that the amount of phosphate covalently bound was found to be a linear function of time, even for long reaction times and up to high degrees of phosphorylation. About 6 mol phosphate per mol of clupeine were incorporated after six days. Since unfractionated clupeine contains on the average 3 to 4 hydroxyl groups per molecule clupeine, one has to elucidate the nature of the excess 2 to 3 phosphates. Acid-Labile and Base-Labile Phosphate The phosphate esters of serine and threonine are known to be acid-stable and base-labile [15]. Table 2 shows data for the different amounts of base-labile and acid-labile phosphate obtained by phosphorylation of unfractionated clupeine. The relative amount of acid-labile phosphate increases with reaction time. It is obvious that in any case acid-labile and baselabile phosphate do not add up to 100 %. Breyer et al. [161 reported phosphorylation of poly(DL-serine) by chlorophosphoric acid and also obtained a considerable amount of acid-labile phosphate. They proposed that acid-labile phosphatases were linked to the a-amino groups (N-a-phosphates), free a-amino groups were obtained after N-0-acyl migrations which involve the peptide carboxyl group and the serine or threonine side-chain OH [17]. A different explanation is the formation of serine and threonine polyphosphate esters; this is supported by the linear rate of phosphorylation and by the time dependency of the formation of acid-labile phosphates and especially by the fact that considerable amounts of phosphate are neither identified as base-labile nor acid-labile (Table 2). Polyphosphates which would
The nature of the acid-labile phosphate was investigated by observing the time course of phosphate liberation in 1 N HC1 at 44 ”C by comparison with different phosphoryl compounds. The results are shown in Fig. 1 which also contains the data reported for phosphorylated poly(m-serine) [16]. N-Phosphocreatine is hydrolysed very quickly indicating that phosphoarginyl bonds should not exist in the present phosphoclupeine; N-a-phosphates, when present, should be split even faster than phosphocreatine. The time courses of phosphate liberation from 5’-ADP, used as a model for polyphosphate esters, phosphorylated clupeine and phosphorylated poly(DL-serine) are very similar, indicating that the acid-labile phosphate is bound similarly in these compounds. The occurrence of the N-0-acyl migration, a prerequisite for the existence of acid-labile N-a-phosphates, can be checked. The ester bonds formed hereby can be split by treatment with 6 N HCl giving rise to peptide fragments with serine as the N-terminal amino acid. A reaction time of 4 h at room temperature was reported to be sufficient to split 50% of possible ester bonds present in clupeine [ 191. Phosphorylated clupeine Z, containing alanine as N-terminal amino acid was incubated for 4 h in 6 N HCl at room temperature and the resulting N-terminal residues were determined by dansylation. Only about one N-terminal serine per 15-N-terminal alanines was found. Assuming the acid-labile phosphate present in the phosphorylated clupeine to be N-a-phosphates, two ester bonds would have been expected (about 2 acid-labile phosphates per molecule were present in the clupeine) thus giving rise to equal amounts of N-terminal serine and alanine after 50% hydrolysis. The rather low value found for serine makes the existence of N-a-phosphates very unlikely. 31
P-NMR Spectroscopy
The nature of the phosphate bond present in phosphoryldted clupeine was elucidated by 31P-NMR spectroscopy. Phosphorylated unfractionated clupeine was used, in order to supply enough material. The
46
Chemical Phosphorylation of Protamines
I
0
I
5
20
15 10 Hydrolysis time ( h )
25
-
Fig. 1. Rate of cleavage of acid-labilephosphatefromphosphorylated clupeine and several related substances. Batches containing total phosphate of 0.7 mM (0,ADP), 0.1 mM (0,N-phosphocreatine) and 0.3 mM (A, phosphorylated unfractionated clupeine containing about 2.5 phosphates per molecule clupeine) were incubated in 1 N HCl at 44 "C and aliquots of 1 ml were withdrawn at different times for determination of inorganic phosphate as described in Experimental Procedure. The data for phosphorylated poly(DL-serine) (0)were taken from the literature [16]. All the data were related to the total amount of acid-labile phosphate obtained after an incubation time of 24 h
++ + 0 I
0 I
0 I
I o=p-o-
j2+3
4 pn2.5 I
I
I
0
I
0Formulae of ( a ) serine monophosphate residue; ( 6 ) serine dbhosphate residue and ( c ) serine oligophosphate residue. The phosphoryl groups exerting equal (or practically equal) chemical shifts in 31PNMR spectroscopy are designated by the same number
assignment of the four signals detected was performed through the pH dependency of their chemical shifts as is schematically shown in Fig. 2A. Two of the signals are shifted to lower field when the pH is raised, whereas the other two are scarcely affected. Hence the former signals must arise from phosphates with two ionizable -OH groups, the latter from phosphates possessing only one -OH group. Fig. 2B indicates the chemical shifts and their pH-dependence of reference compounds such as serine phosphate, ADP and ATP (chemical shifts of ADP and ATP are taken from [20]). A comparison of Fig. 2A and 2 B clearly shows that location and pH-dependence of signal 1 (Fig. 2A) closely resembles that of serine phosphate (Fig. 2B), signal 2 resembles those of P (/?)of ADP and P (y) of ATP, signal 3 those of the P (a) of ADP and ATP, respectively, and signal 4 that of P (fi) of ATP. This strongly supports the conclusion that phosphorylated clupeine contains monophosphates of serine (and threonine) and polyphosphate esters
pH 7.2
I 0
I
+ 10
I +x)
I 30
Chemical shift (ppm)
Fig. 2. Schematic positioning of the I' P resonances ofphosphorylated clupeine and of phosphoserine, N-phosphocreatine, A D P and ATP at different p H values. Concentration of clupeine was 20 mM containing about 3 phosphates per clupeine molecule. Clupeine was dissolved in 6 M urea containing 0.2 mM EDTA to enhance solubility. The number of scans was 9000, the pH was adjusted with HCI or NaOH. (A) "P resonances of phosphorylated clupeine, the numbers refer to the phosphoryl groups illustrated in the formulae, (B) "P resonances of the reference compounds; phosphoserine (-.-.-)and N-phosphocreatine (....-.) were recorded with a 0.1 M solution in 0.2 mM EDTA. the number of scans were about 10 to 20; the spectra of ADP (----) and ATP (-) were taken from the literature [20]; these had been recorded at pH values of 2.2, 6.15 and 6.9 in the case of ADP and of 3.6, 6.15 and 7.15 in the case of ATP. The chemical shifts are related to external 85 phosphoric acid
47
L. Willmitzer and K. G. Wagner
w
I
,
1
I
10
x)
30
40
Fraction number
Fig. 3. Evidence that in clupeine both serine and threonine residues ure phosphoryhted. 40 mg of phosphorylated unfractionated clupeine containing about 2 mol of phosphate per mol of protamine were hydrolysed with 2 N HCI for 8 h at 104 "C (total volume 3 ml), applied onto a small Dowex 50 WX 8 column (H+ form, 1.6 x 28 cm), and eluted with 0.05 N HCI [21]. The fractions (4 ml) were checked for inorganic phosphate (A), organic phosphate (0)and a-amino groups with fluorescamine (0)
of these hydroxy amino acids, the latter constitute the acid-labile phosphates. This conclusion is strengthened by the findings that treatment of phosphorylated clupeine with 1 N HC1 at 44 "Cfor 14 h (cf. Fig. 1) and removal of inorganic phosphate by gel filtration leaves phosphorylated clupeine with a 31P-NMR spectrum, which contains only signal 1 (serine and threonine phosphate) whereas the signals 2 to 4 have vanished. The Phosphorylated Amino Acid Residues
In order to clarify whether, under the reaction conditions employed, both serine and threonine residues are phosphorylated, unfractionated clupeine was phosphorylated, hydrolysed by 2 N HCl (104 "C, 8 h) and lyophyilized. By the method proposed by Schaffer et al. [21] the hydrolysate was applied onto a small Dowex 50 (H' form) column, after elution of the acid residues and phosphate with 0.05 N HCI, the fractions were tested for phosphate and for aamino groups with fluorescamine. The results shown in Fig. 3 indicate a preceding inorganic and two organic phosphate peaks, the latter are fluorescamine positive. Running the column with authenic substances the first fluorescamine positive peak was assigned to phosphoserine and the second to phosphothreonine. These results, further supported by high-voltage paper electrophoresis, show that under the present experimental conditions both serine and threonine residues are phosphorylated.
Cleavage of Pyrophosphate Bonds
Preparation of phosphorylated protamines, containing only monophosphate esters of serine or threonine, requires cleavage of the pyrophosphate bonds present. This was done by treatment with HCI; however, to reduce concomitant cleavage of peptide bonds, the conditions of Fig. 1 were not applied but hydrolysis was performed at room temperature with 6 N HC1. Following the time course of phosphate liberation it was found that the reaction is completed after about 6 h. The occurrence of peptide bond cleavage is manifested by the appearance of new N-terminal amino acids. This was checked with clupeine Z which has alanine as the N-terminal residue, two internal alanines, 3 serines and 21 arginines. Treatment of phosphorylated clupeine Z for 6 h with 6 N HCl and subsequent N-terminal amino acid analysis revealed 1 - 2 serine residues and 2- 3 arginine residues per 15 alanine residues. These results indicate that during acid treatment partial peptide bond cleavage occurred. Chromatographic Fractionation of Phosphorylated Clupeine
Intact phosphorylated clupeine Z was separated from peptide fragments by chromatography on Biogel P4 as shown in Fig. 4. Analysis of the N-terminal amino acids (cf. the legend of Fig. 4) reveals that the first peak emerging consists of about 95 "/, intact clupeine Z. The amount of N-terminal alanine can be
Chemical Phosphorylation of Protarnines
48
-
20
10
30
Fraction number
Fig. 4. Gelfiltrotion ofphosphorylufed clupeine 2 after cleuwge of'rhe p,vropphosphure bonds. As describcd in Experimental Procedure clupeine fraction Z (50 mg), after phosphorylation and cleavage of the pyrophosphate bonds, was fractionated on Biogel P4,in order to separate peptide fragments obtained during acid treatment. The fractions (10 ml) were collected and combined as indicated in the figure and the N-terminal amino acids were determined as described in Experimental Procedure. The following numbers of N-terminal serine and arginine related to 50 N-terminal alanine residues (alanine is the N-terminal residue of intact clupeine Z) were found: fraction I 2 arginine and less than 1 serine; fraction I1 5 argininc and 2 serine; fraction 111 23 arginine and fraction IV 41 arginine and 6 serine residues
0.10
-
. E
0
-._5E 0.05 -% 0
10
20 Fraction number
Fig. 5. Sepurution of phosphorylufed clupeine 2 according to the dtgree qf'phosphor~,lution. Fraction 1 from Fig. 4 (about 30 mg protein) was lyophilized, dissolved in about 10 ml of 0.05 M sodium acetate of pH 5.7 containing 1.2 M NaCI, applied to a Sephadex C25 column (1.6 x 16 cm) and the column was washed with the same buffer. Phosphorylated clupeine Z components were eluted (20 ml/h) by the following stepwise NaCl gradients (containing 0.05 M sodium acetate of pH 5.7): fraction 8 to 18 1.35 M ; fraction 18 to the end 1.5 M NaCI. Fractions of 10 ml were collected and andlysed for organic phosphate (A) and protein content (0).The inserted small figure indicates the number of phosphates per clupeine Z molecule determined for the different fractions as described in Experimental Procedure
taken as a measure of intact clupeine Z, for peptide bond cleavage at the alanine residues can be neglected, as there are only two internal alanines per 31 residues. Furthermore these bonds are more stable than serine peptide bonds [22]. Amino acid analysis of fraction I of Fig. 4 as shown in Table 1 also confirms that it consists mostly of intact clupeine. The reduction of the serine value usually found with acid hydrolysis [23] is further augmented due to the presence of 0phosphoserine residues [24].
Fraction I of Fig. 4 and the respective fractions obtained by phosphorylating clupeine YI were reanalyzed to determine the nature of the phosphate bond. The amount of acid-labile phosphate found did not exceed 5 to 6% of the total phosphate; this should be reduced upon slightly prolonging the acid treatment. Separation of the phosphorylated clupeine Z (fraction I of Fig. 4) according to the degree of phosphorylation was performed by chromatography on Sephadex CM using a stepwise gradient of NaCl as
49
L. Willrnitzer and K. G. Wagner
shown in Fig. 5. The appearance of three distinct phosphate-containing fractions with clupeine Z, which has three serine and no threonine residue, suggests that fractions of singly, doubly and triply phosphorylated clupeine were obtained. This is confirmed by determining the phosphate/clupeine ratio; the results are indicated at the top of Fig. 5. The triply phosphorylated clupeine should accordingly represent a homogeneous fraction of phosphorylated clupeine, while the doubly and singly phosphorylated fractions may represent mixtures of the respective isomers. The phosphorylation of clupeine YI, which contains 3 serine and 2 threonine residues, was performed in a similar way. After chromatographic resolution according to the degree of phosphorylation a minor fraction containing 5 phosphates per molecule and fractions with less phosphate were obtained.
Enzymic Cleavage of Phosphate Bonds To demonstrate that the phosphate esters of phosphorylated clupeine can be cleaved enzymically, phosphorylated fraction YI was treated with alkaline phosphatase from calf mucosa. 1 mg clupeine Y1 containing about 4 phosphates per molecule was incubated with 20 pg phosphatase for 30 min at 30 "C in 1 ml of 5 0 m M Tris, 1 mM dithiothreitol, 0.1 M NaCl and 2 mM MgC1, of pH 7.5. After removal of the proteins as described in Experimental Procedure 2506 of the input phosphate was determined as inorganic phosphate. This findings are consistent with those of Meisler and Langan [ 5 ] , who reported that this alkaline phosphatase releases phosphate from enzymically phosphorylated salmine at a strongly reduced rate relative to low-molecular-weight substrates such as glycerol 2-phosphate.
CONCLUSIONS The present work describes the first investigation
to obtain homogeneous fractions of phosphorylated protamines by a chemical method. After completion of this work Ullman and Perlman published a paper [25] on chemically phosphorylated protamines, which they used as substrates for phosphoprotein phosphalases. However, they started from unfractionated protamine and succeeded in incorporating only about one mole of phosphate per mole of protamine by reaction with inorganic phosphate and trichloroacetonitrile. They did not try to fractionate the mixture of different phosphorylated species obtained. Enzymic phosphorylation of protamine was described by Meisler and Langan [ 5 ] ; starting from unfractionated protamine they also reported about one mol of phosphate incorporated into one mol of protamine by a histone kinase. Phosphorylation in vivo by
hormonal induction of trout testes and extraction of phosphorylated protamines has been mentioned above. Enzymatic phosphorylation certainly circumvents the formation of polyphosphate esters obtained by the chemical methods. However, it is easy to split the pyrophosphate bonds. Thus, the chemical method, as described in the present work, has the following important advantages : possibility of phosphorylating large amounts of protamines, control of the degree of phosphorylation and possibility of obtaining highly and practically fully phosphorylated species. By analytical procedures and by ,lP-NMR spectroscopy the nature of the acid-labile phosphate obtained by chemical phosphorylation was elucidated and found to originate from polyphosphate esters of serine and threonine. A comparison of the rate of cleavage of the acid-labile phosphate obtained by phosphorylation of poly(DL-serine) with chlorophosphoric acid in the work of Breyer et al. [16] with that of phosphorylated protamines of the present work strongly favors the suggestion that phosphorylated poly(m-serine) contained serine polyphosphate esters and not N-a-phosphates; the latter should be hydrolysed very quickly in acid conditions, faster than creatine phosphate (cf. Fig. 1). It is also likely that in the phosphorylated protamine of the work of Ullman and Perlman [25] the acid-labile phosphate found is serine polyphosphate and not phosphoarginine as the authors suggest. The phosphorylation with POCl, in anhydrous trimethylphosphate constitutes a mild procedure for phosphate incorporation into protamine as is demonstrated by the slow rate of this reaction. An essential prerequisite for the success of this procedure was the fact that the hydrophobic counterion capronate was used. Together with the action of the applied POCl, this leads to complete dissolution of protamine. Application of chloride or acetate salts of protamine resulted in turbid suspensions and a failure to obtain markable phosphate incorporation. The circular dichroic spectrum recorded with protamine capronate in trimethylphosphate after addition of POCl, revealed that about 502; of the peptide bonds are in a-helical conformation, a situation which is only obtained by dissolving protamine in a strong a-helix-forming solvent like dichloroethanol [26]. We are very gratcful to D r M.-R. Kula for performing the iimino acid analysis and to Dr V. Wray for measuring thc magnetic rcsonance spectra and for linguistic advice. This work was supported by the Bunrlesniiiilsterium fur Forscllung unri 7 i d i n d o g k (RCT 16) and the Dwische F~~rschun~ge~ntc~in.s~liuf/ (Wa 91 15. 7).
REFERENCES 1. Ando, T., Yamasdki. M. & Suzuki, K. (1973) Prolamine.s,
Springcr-Verlag, Berlin-Heidelberg-New York. 2. Bretzel. G. (1972) Hoppe-Seyler's Z. Physioi. ('hem. 353, 933 943.
50
L. Willmitzer and K. G. Wagner: Chemical Phosphorylation of Protamines
3. Louie, A. J. & Dixon, G. H. (1974) Can. J. Biochem. 52, 536546. 4. Louie, A. J. & Dixon, G. H. (1972) J. Biol. Chem. 247, 79627968. 5 . Meisler, M. H. & Langan, T. A. (1969) J. Biol. Chem. 244, 4961 -4968. 6. Ando, T. & Watanabe, S. (1969) Int. J. Protein Res. I , 221 - 224. 7. Ames, B. N. (1966) Merhods Enzymol. 8, 115-118. 8. Ahmed, K. & Judah, J. D. (1967) Methods Enzymol. 10, 777780. 9. Martin, J. B. & Doty, P. M. (1949) Anal. Chem. 21, 965-967. 10. Zamenhof, S. (1957) Methodr Enzymol. 3, 702. 11. Bernardo, S., Weigele, M., Toome, V., Manhart, K., h i m gruber, W., Bohlen, P., Stein, S. & Udenfriend, S. (1974) Arch. Biochem. Biophys. 163, 390- 399. 12. Spivak, V. A., Levjant, M. I., Katrukha, S. P. & Varshavsky, J. M. (1971) Anal. Biochem. 44, 503-518. 13. Gros, C. & Labouesse, B. (1969) Eur. J. Biochem. 7,463-470. 14. Seiler, N. & Wiechmann, J. (1964) Experientia (Busel) 20, 559- 560.
15. Anderson, L. & Kelley, J. J. (1959) J. Am. Chem. Sor. 81, 2275 - 2216. 16. Breyer, U., Lapidot, Y., Kurtz, J., Bohak, Z. & Katchalski, E. (1967) Monursh. Chem. 98, 1386-1394. 17. Iwai, K. & Ando, T. (1967) Methods Enzymol. 11,263-282. 18. Flynn, R. M., Jones, M. E. & Lipmann, F. (1954)J. Biol. Chem. 211,791-796. 19. Iwai, K. (1960) Nippon Kagaku Zusshi, 81, 1302. 20. Cohn, M. & Hughes, T. R. (1960) J . Biol. Chem. 235, 32503253. 21. Schaffer, N. K., May, S. M. & Summerson, W. H. (1954) J. Biol. Chem. 206,201 - 207. 22. Hill, R. L. (1965) Adv. Protein Chem. 20, 37- 107. 23. Needleman, S. B. (1970) Protein Sequence Determination, pp. 93 -95, Springer-Verlag, Berlin-Heidelberg-New York. 24. Allerton, S. E. & Perlmann, G. E. (1965) J. Biol. Chem. 240, 3892- 3898. 25. Ullman, B. & Perlman, R. L. (1975) Biochem. Biophys. Res. Commun. 63,424-432. 26. Suzuki, K. & Ando, T . (1968) J . Biochem. (Tokyo) 63, 403405.
L. Willmitzer and K. G. Wagner, Gesellschaft fur Molekularbiologische Forschung mbH, D-3301 Stockheim iiber Braunschweig, Mascheroder Weg 1, Federal Republic of Germany
