At pH 6, with 8 m&r MgC12, sodium maleate accelerated the. Mg++ATPase .... CF, at pH 6. Sodium malrate was the most effective, but malonate, itaconate, and.
THE JOURNAL Vol. 247, No. 20,Issue
OF BIOLOGICAL CHEMIGTRY of October 25,~~. 6506-6510,1972
Printedin
Partial
Resolution
Catalyzing XI.
U.S.A.
of the Enzymes
Photophosphorylation
MAGNESIUM-ADENOSINE TRIPHOSPHATASE FACTOR I FROM CHLOROPLASTS”
PROPERTIES
OF HEAT-ACTIVATED
(Received for publication, NATHAN
NELSON,~
HANNAH
NELSON,
AND
EFRAIM
Conditions are reported under which purified coupling factor 1 from spinach chloroplasts, suitably activated, exhibited Mg++-dependent ATPase activity at rates of about 15 pmoles of Pi released per mg of protein per min. The presence of a carboxylic acid and proper MgClz concentrations were the main factors determining the rate of hydrolysis. At pH 6, with 8 m&r MgC12, sodium maleate accelerated the Mg++ATPase activity of coupling factor 1 up to 30-fold. At pH 8, 2. mM MgClz was optimal and bicarbonate was more effective than maleate as an activator. Higher concentrations of MgClz inhibited the reaction. A monospecific antibody against coupling factor 1 inhibited both the Mg++- and the Ca++-dependent ATPase activities. The specificity for triphosphonucleotides was similar for the two reactions. The K,,, of the Mg++ATPase for ATP was 0.1 m&r at pH 6 with Z-(N-morpholino)ethane sulfonic acid buffer and 1.1 mM with maleate buffer. It is concluded that carboxylic acids increase the V,n,, by removing inhibitory products of the reaction. From the mode of inhibition by ADP, existence of at least two active sites acting cooperatively, is proposed.
May 22, 1972)
RACKER
From the Section of Biochemistry and Molecular Biology, Cornell
SUMMARY
COUPLING
University, Ithaca., New York 14850
enzyme was inhibited approximately 50% (1) by Mg++ at 0.3 mM. Since the photophosphorylation and light-induced ATPase activities of chloroplasts (4-6) are dependent on Mg+f, the Ca++ATPase activity of CFI is not directly applicable to the study of the phosphorylating mechanism of the chloroplast. It is the purpose of this paper to describe conditions under which CFI can be induced to catalyze a rapid hydrolysis of ATP dependent on the presence of Mg++ without loss of coupling activity. The cooperative properties of this Mg++ATPase were studied and the regulatory featured of the enzyme are discussed. EXPERIMENTAL
PROCEDURE
Materials Tricine, ATP, ADP, CTP, GTP, UTP, ITP, digitonin, and bovine serum albumin were obtained from Sigma. TPCKtrypsin and soybeau trypsin inhibitor were from Worthington. [y-32P]ATP was obtained by photophosphorylation of ADP in the presence of 32Pi (7). The resulting [s2P]ATP was purified on a Dowex l-Cl column according to the procedure of Cohn and Carter (8). Tricine buffers and Tricine-maleate buffers were prepared by adjusting the pH with NaOH.
Analytical Methods Unlike the mitochondrial coupling factor 1 (Fi),’ the purified coupling factor from chloroplasts (CFJ has no ATPase activity unless properly activated. Treatment by trypsin (I), heat (1, 2), or DTT (3) induced a rapid BTPase activity which was dependent on the presence of Ca+f. Other divalent ions such as Ni+f, Mg++, Mn+f, Co++, and Sr++ at 10 mM were less than 3y0 as effective as Ca++. The Ca++ATPase activity of the * This work was supported by Grant GB-30850X from the National Science Foundation. $ Present address, Department of Biology, Technion, Haifa, Israel. 1 The abbreviations used are: F,, mitochondrial ATPase coupling factor 1; CFI, chloroplast coupling factor 1; DTT, dithiothreitol; Tricine, tris(hydroxymethyl)methylglycine; TPCKtrypsin, trypsin treated with L-l-tosylamido-2-phenylethyl chloromethyl ketone, a specific chymotrypsin inhibitor; MES, 2-(N-morpholino)ethane sulfonic acid; PMS, N-methylphenaxonium methosulfate.
ATPase activity was measured by release of 32Pi from ylabeled ATP as follows. In a final volume of 1 ml, approximately 4 mM [“ZPJATP (50,000 cpm) and 8 InM CaCh or MgClz were incubated at 37” and the reaction was started by addition of 2 to 10 pg of activated CF1. After 10 min the reaction was terminated by addition of 0.1 ml of 30% trichloroacetic acid and 4 ml of 1.2% ammonium molybdate in 1 N HCl. The liberated 32Pi was extracted with 7 ml of isobutyl alcoholbenzene-acetone (9). The extract, 1 ml, was dried and counted in a Nuclear Chicago gas flow counter. The water phase of a zero time sample was counted to give total counts added. Protein concentration was determined according to the method of Lowry et al. (10).
Preparations Heat Activation of CF1-Spinach chloroplast CFr was prepared as previously described by Lien and Racker (11). Only
6506
6507 those fractions w&h a fluorescence ratio at Aaoa:Aa6” of more than I were used. Samples of about 2 mg of CF1 stored at 4” in 50r0 saturated ammonium sulfate were centrifuged at 10,000 x g for 10 min. The precipitate was dissolved in 1 ml of a solution containing 10 mM Tricine, pH 8, and 1 mM EDTA. The enzyme (1 to 2 mg per ml) was heated either directly, 01 after desalting on a &hades G-50 coarse column equilibrated with the same solution. After 4 min at 64” in a medium containing 10 IIIM Tricine (pH 8), 1 ITAM EDTA, 5 mM DTT, 20 mM ATP, and 0.5 to 1% digit,onin, the heated preparation was Purified F1 was obtained as cooled by cold running tap water. previously described (12). Preparation of Stable EDTA-treated Chloroplasts--EDTAt.reated chloroplasts have been reported to be unstable (13). We found now that in the presence of high concentrations of bovine serum albumin, t,hese particles can be preserved at, -70”. Spinach chloroplasts were prepared as previously described (14) except t,hat 1 mg of bovine serum albumin per ml was used and swelling of the chloroplasts was performed with 10 InM Tricine (~1-1 8) at 4”. The suspension was centrifuged at 20,000 X g for 10 min and the resulting pellet was suspended at a chlorophyll concentration of 2 to 3 mg per ml in a medium containing 0.4 M sucrose, 0.01 M NaCl, 0.01 M Tricine (pH 8), and 1 7. bovine serum albumin. These chloroplasts were diluted t,o a concentration of 0.1 mg per ml with 0.75 m&I EDTA (pH 7.2). After 10 min at 4”, the chloroplasts were centrifuged at 35,000 X g for 15 min and suspended in the sa,me medium described above at, a chlorophyll concentration of about 2 mg per ml. The EDTX particles were divided to 0.5~1x11 portions frozen immediately and thawed just before use. These preparations were stable at -70” for months and gave rise to photophosphorylat,ion in the absence of CF, of 2 to 4 pmoles of ATP per mg of chlorophyll per hour, which was increased to 60 t,o 120 pmoles on addition of CFI.
RESULTS
Actzvation of Mg++-dependent ATPase Activity of CF1-When CFI was heat-activated and tested at pH 8.0 (2), as described previously, little Mg++-dependent ATPase activity could be detected (about 3% of the ratd with Ca+f). However, when t.he enzyme was tested at pH 6.0 in the presence of maleate buffer, considerable Mg++ATPase activity was observed. An exploration of this observation revealed that the activity in the presence of 8 m&I MgClp was dependent both on the presence of male&e and the lower pH. As shown ia Fig. 1, increasing concentrations of maleate-stimulated Mg++ATPase activity up to 15-fold whereas Ca++ATPasc activity was inhibited so that above 60 nlM maleate the activity with T\lg++ was actually higher than with Ca++. With 8 mM MgCl, in the reaction mixture, 6 was the optimal pH for Mg+f-4TPase activity of heat-activated CF1 (Fig. 2). Similar results were obtained with trypsill-activated CFI, while DTT activat,ion gave rise to low iVfg+i-ATPase activity peaking
at a higher
$1.
Additional
studies
showed
that
the
opt.imal pH for t,he Mg+fATPase reaction is markedly affected by the MgCl~ concentration. As can be seen in Fig. 3, even higher Mg++ATPase activities can be obtained at pH 8 with the appropriate XIgC$ concentration. The Mg++ATPase at pH 8 has a sharp optimum around 2 rnM MgCl, while higher concentrations are inhibitory. The activity at pH 6 shows much less sensitivity toward high MgCl, concentrations. But, even in 8 mM MgClz at pH 8.0, a specific activity obtained provided sodium malcate was present mixture.
Table I sumrnarizes
of 3 to 4 was in the reaction
the effect of various organic
acids on the
Mgf+ATl’ase activity of heat-activabed CF, at pH 6. Sodium malrate was the most effective, but malonate, itaconate, and phthalate also accelerated the activity by more than IO-fold. At 1~1-1 8, with 2 mM MgC12, sodium bicarbonate was more
Heat -ATPose
DTT - ATPose 3 I
I 20 Sodium
I 40 Moleate
I 60 (mM)
I 80
I 5.5
I 6
I 6.5
FIG. 1 (Zefft). Effect of sodium maleate on Mg++ATPase activity at pH 6. The reaction mixture contained in‘a final volume of -1 ml : 20 umoles of MES (uH 6). 4 umoles of ATP. 8 wmoles of CaClz or Mgbl~, and 6 rg of heatlactivated CFI (Ace’ “Experimental Procedure”). FIG. 2 (center). Effect of pH on Mg++ATPase activity. Heatactivated CF1 was obtained as described under “Experimental Procedure.” Activation of CF, by 1>TT was performed by incubating the enzvme with 10 rnM Tricine (nH 8). 1 m&l EDTA. 50 mM DTT: and 10 -rnM ATP at room temperatire for 4 hours. The Ca++ATPase activities at pH 8 were 40 and 28 pmoles of P, per
PH
/
I 7
I 75
I Mg
Cl,
(mM)
mg of protein per min for the heatand DTT-activated CFI, respectively. The assay mixture for Mg”ATPase contained in a final volume of 1 ml: 20 pmoles of Tricine-maleate at the indicated pH values, 4 pmoles of ATP, 8 pmoles of MgCls, and 10 pg of CF1 activated by heat. or by DTT. FIG. 3 (right). Effect of MgClz concent,ration on Mg++ATPase mixture contained activity of CF1 at pH 6 and 8. The reaction in a final volume oi 1 ml: 60 pmoles of Tricine-maleate (p1-I 6 or 8). 4 wmoles of ATP and M&In as specified, and 5 LIE of heata%ivat,ed CFI.
6508 effective than maleate. .1s can be seen in Fig. the reaction by 3.5fold, giving rise to an activity 16 I.cmoles of ATP hydrolyzed per mg of protein Sensitivity
io SpeciJc
Inhibitors
and
SpeciJcity
4, it accelerated of about 14 to per min. toward
Nucleo-
Mg++ATPase arnoles
Sane Tartrate.
Oxalate Sllccinate.
........
Isocit,rate.
........ ........
Glycolate. Fumarate Ithaconate.
Phthalate. Malate. Malonate Malcate
_. _.
I
I
I
UTP
was
activity
only
about
3%
that
of’ XTP
and
no
hmole
in
yielded
LEVIES buffer,
a K,
of Pi per
pH
for ATP
mg of protein
6, at
of 0.11 per
various i\ddition
min.
hTP
and a V,,,,
InM
of 60
con-
of mM
sodiurn maleate increased the V,,,,, to 14 pmoles of Pi per mg of protein per min and the K, for ATP to 1.1 mM (Fig. 5). This shows clearly that the acceleration of the lZT1’ase activity by maleate is not caused by a greater affinity of A’l’l’ for CF1. .1t pH 8, with sodium maleate in the reaction mixture the K, for AT1 was 1.8 mM. It was previously observed (1) that XDI’ strongly inhibited the Ca++ATPasc activity of CF1. As can be seen from Fig. 6, Tarrm
II
EJect of aniibocly on Ca++and Mg++Arl’Pase nclivities of CFI The reaction mixture for measuring Ca++AT!‘ase contained: 30 pmoles of Tricine, pH 8, 4 pmoles of ATP, and 8 pmoles of CaCln. For measuring Mg++ATPase at pH 8: 60 pmoles of Tritine-maleate (pH S), 4 pmoles of ATP, and 2 pmoles of MgC12. For measuring Mg++ATPase at pH 6: 60 pmoles of Tricine-mnleate (pH 6)) 4 pmoles of ATP, and 8 rmoles of MgCln.
activity protein/?&
0.3 1.6 1.8 2.5 2.1 2.8 2.3 4.0 3.4 2.0 7.1 7.6
........ ........ ........ ........
.,
Pdmg
ATPase 0.45
I’: I
acid
was detected
centrations
Eflecl of c~arhoxylic acids on Mg++ATPase activity/ of CF, at pH 6.0 The reaction mixture contained in a final volume of 1 ml: 20 @moles of MI‘S (pH 6), 8 pmoles of MgCl,, 4 pmoles of ATP, and 2.8 pg of heat-activated CF1. 40 III&I organic
with
activity
with CTP as the substrate. Kinetic Properties of ATPase-‘l’o elucidate the mode of action of organic acids on the A’l’Pase activity of CF1, some kinetic A Lineweaver-I%urk plot for XIg++parameters were examined.
tides-When monospecific antibody against CF, (15) was added to the reaction mixture described in Table II, Ca++hTl’ase and Mg++dTPasc (pH 8) as well as 1Llg++ATPase (pII 6) activities were similarly inhibited. p;o-9 also inhibited both l\1g++- and Ca+f-kTl’ase. IIowever, the Ca++ATPase was more sensitive, as can be seen from Table III. The specificity of Mg+fA4TPase toward various nucleotides was similar to that of the Ca++hTPase (1). -kTP was hydrolyzed about 6 times more rapidly than GTP or ITP. The T,zur,
activity
Additions
pmoles
Control Control
(1.7 @g CF1) + 10 ~1 control
Control
+ 20 pl control serum.
Control Control
+ +
10 ~1 anti-CF1. 20 ~1 anti-CFI
serum. .,
_.
47.4 43.2 36.4 7.3 3.0
Pi
l.eleased/mg puotcin/mi?z G.7 13.6 17.2 16.7 17.2 16.0 4.2 2.8 2.9 / 1.3
I
I
I
I
I
I
,
I
25
I
I
I
100
2 I/S
FIG. 4 (left). Effect of sodium bicarbonate on Mg’+ATPase activity at pH 8. The reaction mixture contained in a final volume of 1 ml: 40 pmoles of Tris-Cl (pH S), 4 pmoles of ATP, 2 pmoles of MgCle, sodium bicarbonat,e as specified, and 6 rg of heat-activated CF,. FIG. 5 (center). Effect of sodillm maleate on the K, for ATP and the V,,,,. of Mg++ATPase. MES, the reaction mixture contained in a final volurne of 1 ml : 20 rmoles of MRS (pH 6)) 8 pmoles of MgCl,, ATP as specified, and 3.5 rg of heat,-activated CF,. Maleate, as above but M15S buffer was omitted and 60 pmoles of Tricine-malcnte (pH 6) were included.
(mM ATP)
mM
ADP
FIG. 6 (right). Inhibition of ATPase activities of CF, by ADP. The reaction mixture for Ca++ATPase contained in a final volume of 1 ml: GO pmoles of Tricine-maleate (pH 8), 4 rrnoles of ATP, 8 pmoles of CaC12, and 3.5 rg of heat-activated CF,. The reaction mixture for Mg++ATPase at pH 8 was the same as for Cat+ ATPase except that CaClz was replaced by 2 pmoles of MgCl,. For Mg++ATPase activity at pH 6, Tricine-maleate, pH 6, and 8 fitmoles of MgC12 were used. The activities without ADP were 35 Fmoles of Pi per mg of protein per min for Ca++ATPase, 10 for Mg++ATPase at pH 8, and 7.7 for Mg++ATPase at pH 6.
6509 ADP also inhibited the hlg++-ATPase activity. In the presence of 4 mM ATP 50y0 inhibition was obtained at the following 0.4 MM for concent.rations of ADP: 0.7 IIIM for Ca++hTPase; Mg++ATl’ase at pH 8, and 0.3 KIM for Mg++ATPasc at pH 6. Hammes and Hilborn (16) reported that in the case of the mitochondrial ATPase, ADP is a competitive inhibitor of ATP. As can be seen from Fig. 7, IDP inhibition of the M&f-ATPase activity of CF1 did not follow the same pattern; both the affinity Moreover, ADP toward ATP and the V,,,, were modified. changed the saturation curve of ATP from a hyperbolic to a sigmoid shape. The apparent reaction order changes from 1.0 in the absence of ADP to 2.3 in t,he presence of ADP (Fig. 8). The same phenomenon was observed for Ca++hTPase and Mg++Al’Pase activities at pH 6. In agreement with the findings of Hammes and Hilborn (16), we found that the Vlllax of the mitochondrial ATPase was not modified by ADP, but the reaction order as determined by Hill plots changed from 0.9 to
Coupling Activity of ,Mg++ATPase-As shown in Table IV, CFI with Mg++ATPase activated by heat in the presence of digitonin retained its coupling activity. The mode of action of digitonin is now under active investigation. DISCUsSION
Previous studies of the mechanism of action of CF1 in its relation to photophosphorylation were complicated by the fact that after activation of the AT’Pasc activity by either trypsin or -
I
I
O-
1.4. :: T.\ISLE EJect
of
Die-9
on CCL++- and
>
III Mg++ATPase
ac/ivities
of CF,
The experiments were performed ns described in Table II except that the bufl’er for measuring Ca++ATPase was 60 pmoles of Tricine-maleate,
Control. Control Control Corit,rol Control
pH
;: E >
-l-
-ii?
+06mM
8.
Additions
ca+ +ATIke
....................... + .i pg 1X0-9. + 10 pg l)io~9. + 50 pg 1X0-9 + 100 pg Dio-9.
27.3 14.1 8.6 1.9 0
........ ....... ........ .......
Mg+‘ATPase (PH 8)
Mg++ATPase (ptl 6)
l.j..j
13.3
14.1 11.1 5 5
10.0 6.4 3.1
-2-
-3-
J
-I log
FIG. mental
s(mM
8. Hill plot of effect of AIlI’ conditions were as described T.\ISLI: Reconstilution
ADP
ATP)
on Mg.+ATPase. in Fig. 7.
Rxperi-
IV
of c~yclic photophospho~!/lalion particles lq h,eat-treated CF1
in ELI TA
CF, was activated by heating in the presence of digitonin as described lmder “Experimental Procedure.” CF, or heat-treated CFI was added to EDTA particles (20 fig of chlorophyll) as indicated in t,he table. The reaction mixt,urc contained in a tot,al volume of 0.3 ml : 20 pmoles of Tricine-N&H (pH 8)) 5 pmoles of NaCl, and 120 pmoles of sucrose. After adding 5 ~1 of 1 M MgCl, the mixtlwes were inc~tbated at 0”. Aft,er 10 min, O&ml samples The assay mixwere assayed for cyclic photophosphorylation. ture for photophosphorylation contained in a final volume of 3 ml : 50 pmoles of Tricine (pH 8)) 00 pmoles of NaCl, 20 pmoles of MgCl2, 10 pmoles of Pi, 3 pmoles of ADP, 0.09 pmole of PMS, and approximately lo6 cpm of 32Pi. After illumination by white light (10” ergs per cm2 per s) for 60 s, the reaction was terminated by xddition of 0.3 ml of 300/;, trichloroncetic acid. Incorporation of “21’X was measured according to the method of Avron (9). Additions
ATP
phosphorylation
~moles A 7‘Plmy chlorophyll/hr
(mM)
FIG. 7. Allosteric inhibition by ADP of Mg++ATPase activity of CF1. The reaction mixture contained in a fmal volume of 1 ml : 60 pmoles of Tricine-maleate (pH 8)) 2 ,umoles of MgCL, ATP as specified, and 2 pg of heat-activated CF,. Values of 20 pmoles of P, per mg of protein per min for the control and 10 in the presence of 0.6 rnM ADP were obtained for the V,,;,, of the reactions by extrapolation to infinite ATP concentration in the LineweaverBurk plot.
Cyclic
EDTA EDTA ICDTA EDTA EDTA I+X~TA lj:l>TA
particles.. + 64 /.~g + 96 I1g + 64 pg + 96 pg + 64 pg + 96 pg
CF,.. _. ., CF,. ._ heated CF,. ., ._ ._ ._ heated CFl. (digitonin) heated CF,. (digitonin) heated CF,.
2.4 115 138 7.9 10.3 85.3 99.8
6510 heat, the protein no longer served as a coupling factor (1, 2). Even mild activation by dithiothreitol yielded a Cn++-drpendrnt ~ZTPase activity (3) while membrane-bouud CF1 (3, 6) catalyzed in the light, a ?\Ig ++-dependent ;\‘l’l’ase activity (4, 5) as well as l)llotol)tlosphorgl~~tiol~. St,raub and Lynn (17) observed act,iv:rt,ion of a M g++-dependent ~l’l’l’ase when illuminated chlorol~l:~xts wcrc treated wit,h TPCK-trypsin. A 3-fold increase in 1Ig+.\Tl’:~se activity after treatment of DT’r-activated -YI’Pasc with TlY’K-trypsin was reported by Lien and Kacker (18), which was further iucreased about 3-fold by addit,ion of C’F1-deplctctl subchloroplast, particles. In all these studies, the highest sllccific activity of M g++-dependcxt Al’l’ase was less than 1 ,~nlole of ;\Tl’ cleaved per mg of CF1 protein per min, while the soluble preparations described in this paper had specific actiritics exceeding 15 without loss of coupling activit,y, provided digitonin was present during heat activat,ion. Without digitonin I\Ig++lZ’l’I’ase was also activated, but coupling act>ivit,y was lost. Sexual aspects of this activation are of interest. Heat trentmerit in the presence of digitonin results in an alteration of l)rotcill st ructurc which is under investigation. The ot,her main Factors contributing to high YIg++llTPase activity are the presence of an organic acid or bicarbonate, and the choice of a suitable 1Ig++ and H+ concentration during assay. The kinetic data suggest that the acceleration of the Mg++ATPase b\ mnleatc is due to acceleration of the turnover of the enzyme, b> increasing t,he rate of either the catalytic step or the product relcasc step, whichever is rate-limiting. Jagcndorf altd Iiribe (19) showed that organic acids markedly accelrratcd ;Yl’P formatiou in chloroplasts in an acid-base l,raiisil~ion. Mthough the major effect of the organic acids was bcliercd to bc to supply an internal buffer capacity to the chloroplasts, it might be that they also play a role in the activation of the CF1 on the chloroplast membrane. Wang et al. (20) reccnt,ly dcmolxtrated that maleate or succinate accelerated -YI’I-’ formation in the dark after a brief illumination of chloroplasts. In thesr experiments succinate, which has better buffering capacit,y in the range of experimental condit,ions employed, should have been more effective than maleate. The fact that it was not suggests that the stimulation might be partly due to an activation of CFI. 13atra and Jagendorf (21) studied the effect of bicarbonate on I)llotol)llos,lior~lation and observed acceleration of photophosphorylation whiln tile yield of the high energy intermediat,e (X,) decreased. ;1t low 1~1-1the step catalyzed by CFi might bc rate-limiting. Thus the acceleration of ATPax activity and of :Yl’I’ formation by bicarbonate may be related. The effect of LII)l’ 011 the activity of Mg++hTl’ase was at least partially cooperative in nature as indicated by the kinetic behavior of the system. Since CF1 contains five diffcrcnt subunits (22) conformational changes may result in different interactions between t.he subuuits. Energy-dependent conformxtional changes of membrane-bound CF1 were demonstrated by Ryrie arid Jagendorf (23). It will be of interest to study possible
similarities between the two effects. The availability of antibodies against each of the five subunits of CF,2 (15) may permit a niorc direct al)l)roacli to this problem. It is very likely that the inhibition of the AYYI’ase activities of both F, al,d CF, by *1I)I’ has a ])hysiological significance in controllilig the brcakdowll of ATI’. It will bc of interest, thcrcfore, to d&ermine which of the subunits int,eract with X)1’ at the active site and which at, the cooperative site. The isolation of individual CF1 subunits should make a direct approach to this problem feasible. There is a remarkable similarity between the properties of mitochondrial F1 and chloroplast CFI in amino acid composition, subunit st,ructure, cold lability, and iuhibition by X111’. Xn apparent diference in their properties has disappeared mit,h the preparation of an active 1\Ig++.1TPase from chloroplasts with coupling activity. V.~MIKJT.IS, V. K., MU 2. 3.
2(X0-2GG7 F.IIU~OK, iV1&.1rtw,
4.
PETX~X,
F., li.
I~XI~ER,
16. (lQ65)
J. Riol.
21~1) IZKKER, E. (1970) Biochemistr?y I
