Control of Phosphorylase Activity in a Muscle Glycogen Particle

SUMMARY This study describes a new type of inhibition of phosphorylase phosphatase that strongly suggests that the activity of this enzyme is modulated during regulation of the phosphorylase system. Phosphorylase phosphatase included in a muscle protein-glycogen complex along with phosphorylase, phosphorylase kinase, and other enzymes of glycogen metabo-lism undergoes a reversible inhibition when phosphorylase is activated. This inhibition requires free Ca2f in addition to Mg-ATP, i.e. the same conditions that trigger phosphorylase activation; it is observed only in the intact protein-glycogen complex. It is not caused by AMP or IMP generated from the rapid breakdown of ATP, two nucleotides known to inhibit the phosphorylase phosphatase reaction by binding to the substrate phosphorylase u. (a) Addition of AMP or IMP to the particulate system is without effect; (b) phosphatase activity returns to normal at the end of an activation cycle of phosphorylase when concentration of IMP is maximum; and (c) the same inhibition pattern is observed


EDMOND H. FISCHER]]
From the Department of Biochemistry, University of Washington, Seattle, Washington 98105 SUMMARY This study describes a new type of inhibition of phosphorylase phosphatase that strongly suggests that the activity of this enzyme is modulated during regulation of the phosphorylase system.
Phosphorylase phosphatase included in a muscle protein-glycogen complex along with phosphorylase, phosphorylase kinase, and other enzymes of glycogen metabolism undergoes a reversible inhibition when phosphorylase is activated.
This inhibition requires free Ca2f in addition to Mg-ATP, i.e. the same conditions that trigger phosphorylase activation; it is observed only in the intact protein-glycogen complex.
It is not caused by AMP or IMP generated from the rapid breakdown of ATP, two nucleotides known to inhibit the phosphorylase phosphatase reaction by binding to the substrate phosphorylase u. (a) Addition of AMP or IMP to the particulate system is without effect; (b) phosphatase activity returns to normal at the end of an activation cycle of phosphorylase when concentration of IMP is maximum; and (c) the same inhibition pattern is observed when a phosphopeptide derived from phosphorylase a is used as substrate, a reaction not affected by AMP or IMP.
Digestion of glycogen in the complex by a-amylase abolishes the Cazt-dependent inhibition of phosphorylase phosphatase, but the activity of this enzyme is still predominantly unaffected by nucleotides.
On the other hand, dissociation of the protein-glycogen complex by dilution brings about both a loss of the Ca2f-dependent inhibition of the phosphatase and the reappearance of its sensitivity to AMP or IMP. modified, e.g. by phosphorylation; rather, it is proposed that this inhibition results from the interaction of phosphorylase phosphatase with some of the other components of the phosphorylase system.
The first paper of this series described the isolation and chnracterization of a protein-glycogen complex from rabbit muscle (1) and it was proposed that this complex is a structural and functional unit of the cell. In the second nrt,icle (2) some aspects of phosphorylase regulation were described, especially with respect to phosphor\-last kinnse nct,ivation. When Ca2f and Mg-ATP were added to this probria-glycogen complex, an immediate activation of phosphorylnse was observed, re.qulting from a phosphorylase b to a colwersioll, followed by a rapid return to the original inactive b form. The over-all reaction was referred to as t,he LYlash activation" of phosphorylase. Previous evidence obtained both with the intact animal and on purified systems all indicntcd that control of phosphorylnse activity occurred through activation of phosphorylase kinase, not inhibition of phosphoryl:rsc phosphatahe (3). Extensive studies on purified phoaphorylwe phosphatase revealed few clues as to its possible involvement in a regulatory process as found for phosphorylast kinnsc or glycogen synthct:lse, aside from its known inhibition by AMP or IMP and its slight activation by methylxnnthines, glucose, or glucose-&P (3)(4)(5)(6)(7). Purified phosphorylase phosphntase is hardly affected, if at all, by cyclic 3') 5'.hMP, various divalent metal ions, or chelating agents such as El1TA, ethylene glycol bis(P-aminoethyl ether).N,X'tetraacetic acid, o-pllcanllthroliiie, etc.; at i;ome stage of its purification, it is activated 2-to 3-fold by limited tryptic attack, but ilot by the Ca2+-dependent kinnse activating factor (7). On the other hand, some changes in activity under the influence of Mg2+, ATP, and cyclic 3',5'-AMP have been reported in other systems including bovine adrenal cortex (8,9), pigeon breast muscle (lo), and dog liver (11). So similar effects hare been observed with the rabbit muscle enzyme.
The present article presents evidence that in the proteinglycogen complex, but not in the disrupted system, phosphorylase plrosphatase displays enzymatic characteristics different from those of the purified enzyme and that its activity can be modulated in a may suggesting that it actively participates in the regulation of glycogen phosphorylase. This is an Open Access article under the CC BY license.

MATERIALS
AKD AIETIIODd applied and the entire pool collected. This fraction was brought ~brterials and methods not listed in this paper have been to 455; saturation with solid ammonium sulfate at 0" and cende,qcribetl in the two preceding articles (1,2). trifuged after 2 to 3 hours. The precipitate n-as resuspended in The plrosphorylasc a labeled n-it.11 32P was prepared according a small volume (approximately 10 ml) of the Tris-mercaptoto I'i-clicr et al. (12) (specific radioactivity approximately 4 X ethanol buffer and dialyzed against the same buffer. The solu-10" p1)m l)er mg). tion obtained (approximately 15 111 g of protein per ml) was stored at 0"; the specific activity of the l~l~ospl~ntasc was npproxi-Phosphoqlase Phosphatase Assays mately 180 units per mg of protein and represented a purification  of V was linear nitli time !vlrcil less t,hnn 505; of the total pro-pAI.
'I'he pep-comples during an actirat,ion cycle of phosl~lioryhlse, samples tide fraction was added to the enzyme solution in the same vol-n-tre wiiiomtl al various times md 0.1 volume of a solutioii of ume ratio (100 ~1 per ml) as used in Method 2 and at a final con-"?l'-labeled phosphorylasc a of high specific rntlio:ict,ivil,y \vns centration of 30 nmoles of 32P per ml. The reaction n-as stopped added. From these samples, three aliquots wcrc rcmovc:tl on by nddirig O&ml samples of t'his reaction mixt.ure to tubes con-time ant1 measured for release oi "Pi, as tlescribctl undrr "Al:itetaining 0.5 ml of a. slurry of Dowex 50-B (II+) in water (0.25 g rials alit1 l\Ictl~ods" (Method 2) from which tlic initial late of per ml) l~lus 0.4 ml of 105; trichloracetic acid. Radioactivity l~ho~phataae activity was calculntcd. If pliospliorylase phosrelease was measured on the supernatant obtained after centrifu-phatase activity is follow-ed in this fashion, a liattern rmcr'ges gation.
Linear initial rates Jvere observed only Khen less than that is illustrated in Fig. 1. One obscrvcs an immctlintc ap-2076 of the total radioactivity was released. prosimately SOCz inhibition of the phosphatasc followctl l)y re-Partially purified 1ilrosphoryl:rsc plrosl~h:rt;~sc Kas 1)rc1):uCd by activatioii of the enzyme at the end of the flash activation. 'l'lris a moclifiration of tlic procedme 0T lIurc1 (i).
The precipitate inhibition is mostly dependem on the presence of free Ca2+ siirce obtained by centrifugntion at 80,000 X 9 (30.pellet) was prc-in the absence of this metal ion n-hci,c no phosphorylase nctiva-1~:1red according to the method of Krebs et al. (12) n-ith the ex-tion occurs, no more than a 20 to 259i;, inhibition of tlrc phoscrl)tioris that 20 nl>r Tris-HCl, 50 m&r P-mercal)toeth:~rlol buffer, phatase is observed. filasimum inliibition occurs at the peak pTI 7.5, was used and the suspension was treated for 10 min at of the flask activation whtre maximum phosphorylnse activity 30" viit,li IO ~1 per ml of n crystalline sus1)ension (approximately is produced. 8 mg per ml) of salivary aniylasc.
'I'hc amylase-treated sus-It lx-as of intcrest to detcrminc whether the observed cffccts pen.Aon was diluted with an equal volume of the same buffer n-ere dependent upon the integrity of tlic protriii-glycogcii cornand centrifuged at 140,000 x g for 3 hours. The supernatant ples or ~vhether they could be i~tl)i~odu(~rd with pmifictl C~Z~IIICR. was then adsorbed on a column of Whatman DE-52 (approxi-If l~hosl~l~orylase pliosphatase in the 30.pellet susl)(~nsioii (2) is mately 60 ml of packed cellulose bed per 600 g of muscle) equili-liberated from the r)rotein-gl!-cog(~l~ romples to whicsh it is bound brated Tvith the same buffer containing 70 mar Sa('l. l'he col-by dilutioii of the system, itx ruzymatjic behavior is iiidced umn was mashed with the equilibration buffer until the A280 of different. This is exemplified in Fig. 2 The same inhibition pattern is observed (Fig. 3)   regulatory enzymes are also present in the proteinglycogen complex.
To examine these two possibilities, inhibition of phosphatase was first studied in a proteinglycogen complex which had under-gone dissociation by dilution. Table I shows the effect of nucleotides other than :yI'P on the activity of the enzyme before and after dissociation.
In the latter instance, and as espected, both IMP and IlLIP inhibited the pl~ospl~ntnsc reaction; this inhibition was previously shown to result from an interaction of the nucleotides with the substrate, phosphorylnse a, rather than Ivith the phosphatase itself (4); iii values of approximately 5 x lo-" 31 were found for AMP and approxim:itely 1 x 1OW iv1 for INP (7).
In the diluted 30-pellet suspension, although one Pees a substantial inhibition of the phosphatnse, the extent of inhibition is not as large as predicted by the above inhibition constants. By contrast, in the undiluted system, containing the intact protein-glycogen complex, essentially no inhibition was caused by .UIP or IMP (added together with Ca2+ and RIg2+) and the slight effect produced by .1DP could be ascribed to the formation of *YIP under the influence of myokinase also present. These data indicate that the reversible inhibition of phosphorylase phosphatase observed durin g flask activation cannot be simply ascribed to a transient production of QIP or IMP from -1TP. This conclusion was confirmed by the use as substrate of a phosphopeptide derived from phosphoryla~e a, as will be described in a later section.
Correlation between Phosphatuse Inhibition and Xucleofides Produced during Phosphorylase Activation--The :~l>ovc conclusions n-cre confirmed by determining the nucleotides generated from .1TP during phoaphorglase activation in t.he disrupted and intact prot'ein-glycogen complex. '1'0 this effect, [8-i4C-,\TP (1 PCi per pmole) was added to an a-am3-l:lse-tre3tcd supernatant solution and to both a diluted and undiluted 30.pellet suspension; at various times, samples were removed and analyzed by thiu lager chromatography as described u~ldcr "hiaterials and Methods." Results arc illustrated in lcigs. 4 alid 5, respectively.
In the dissociated systems, recol-erics of all four nuclcotides varied from 82 to lOO%, as coml):nctl to only 49 and 73:; in the undiluted suspensiou, becau>e of the large :unount of precil)it:Lte formed during deproteinization; ho~cvcr, no more than 2' c radioactivity could be detected on the tllin layer plate in areas not represented by dTP, ADI', UIP, or 1311~.
Rrcaktlown of -IT1 was much slower in the amylase-trcatcd supernatant (Fig. 4) since the -1TPnse associated wit11 the elements of the sarcoplasmic ret.iculum had been removed.
In all preparations, myokinase and LhIP deaminase were highly active as indicated by the low steady st,ate levels of ADP and -IMP; in all instances, there was ample IMP to completely inhibit the phosphatase as observed in the "dissociated" systems. Yet, in the particulate system, t'here was no correlation between IMP accumulation and enzyme inhibition. Fig. 1 shows that the enzyme is maximally inhibited within 30 set but returns to normal after 4 to 5 min when the concentration of IRIP is maximum.
Phosphorylase Phosphafase Inhibition with Phosphopeplide Derived from Phosphorylase a as Substrate--The nl,o\e data indicate that in the proteinglycogen complex (as opposed to the dissociation system), inhibition of phosphorylase phosphatase results from a decrease in the activity of tliis enzyme, not from a reduced susceptibility of the substrate to enzymatic tlrpho~phorylation as afforded by UIP. Nest coiiviilciilg cvitlcilce was obtained by the use as substrate of an 80.amino :rclitl, 321'. labeled phosphopeptide derived from pl~osl~l~oryla~ CL 1)y ('NRr cleavage (13) . This phosphopept~ide is del~l~o~~~llo~~~latecl at 25% of the rate at which phosphorylase a is attacked; horn-ever, the former reaction is totally unaffected by 1 rnhf AMP or IMP. Table II clearly indicates that in the intact system (undiluted 30-pellet suspension) essentially the same inhibition pattern was observed when the phosphopeptide was used as substrate (mnximum inhibition with Ca*+ and Mg-ATP, little with Mg-XTP alone, and none with Mg-IMP).
As in Fig. 1, the inhibition disappeared within 5 min. By contrast, in the dissociated system essentially no reduction in activity occurred confirming once more that the observed inhibitions were caused by the production of AMP or IMP.
l?vidence against Covalent Modificafion of Phosphorylase Phosphatase during Inhibition-The most likely mechanism for the inhibition described above might entail a covalent modification (e.g. phosphorylation) of the protein.
If this were the case, one would expect that the phosphatase would remain inhibited even after dissocint.ion from the other components of the system by a process such as high dilution.
To examine &is possibility, the undiluted 30-pellet suspension was incubated with Ca2+ and Mg-ATP; after 30 set when phosphatnse inhibition wasmaximum (75%) the material was rapidly diluted loo-fold into 50 mu sodium glycerophosphatc, 1 mar EDTA buffer, pH 6.8, and assayed. No inhibition remained as compared to a control witllout ,ihIP. It can be concluded, therefore, that unless substitution is rapidly reversed (e.g. dephosphorylation) inhibition of phosphorylase phosphatase in the intact protein-glycogen complex does not result from a covalent modification of the protein.
13Ject of Glycogen Degradation on Phosphorylase Phosphafase Activity in Protein-Glycogen Complex-As shown above, dissociation of the 30-pellet suspension by dilution drastically affects the conditions under which phosphorylase phosphatase is inhibited.
It was, therefore, of interest to determine the role of glycogen in maintaining the type of reversible inhibition found in the original protein-glycogen complex.
To this effect, the 30-pellet suspension was digested with cu-amylase and phosphatase inhibition was studied as a function of glycogen removal and restoration.
As seen in Table IIIA, removal of glycogen resulted in a marked loss in the inhibition caused by Ca2+ and Mg-ATP, and an increase in t,he inhibition caused by IMP. However, this inhibition is still much lower than that observed in the complex disrupted by dilution (Table IIIB).
Readdition of glycogen to the undiluted amylase-treated suspension did not restore the original inllibition pattern (Table 111s). In contrast, hydrolysis of glycogcn reduces the flash activation of phosphorylase by al)proximately 2-fold and this effect, is completely reversed by renddition of glycogen (2). Therefore, phosphorylase kinase bchnvcs differently than phosphorylnse phosphxtase in that its original enzymatic characteristics are restored by renddition of glycogc11.
Rcaddition of glycogen (291) to the proteill-glycogen complex disrupted by dilution did not reverse the IMP inhibition of phosphorylase phosphatasc (Table IIIB).
Tllis indicates that appcnrance of the IMP sensitirity upon dilution is not simply callsed by a lowering of the glycogen concentration, but perhaps depends on protein-protein interaction.
Comparison between Phosphorykse Phosphatase Activity on Endogcnolls and on Exogenous Phosphorylase a-The data prescntctl thus far indicate that phosphorylase phosphatase is associ:ttcd with a protein-glycogen complex in which its enzymatic are different from those of t,he dissociated enzyme. One must, therefore, ask what effect this particulate system has on the interaction of phosphorylnse phosphatase with soluble phosphorylase a :IS compared to endogenous phosphorylnse a present in the protein-glycogen comples.
In the 30.pellet suspension, one can measure phosphorylase phosphntase activity either by addition of 32P-labeled phosphor~~l:lse a (qee "Materials and Met.hods") or loss of endogenous phosphorylase a activity produced in the 30.pellet suspension during fla>k activation.
This lnt,ter procedure gives an nccuratc measure of phosphorylnse phosplintasc activity only after ;2TP has been depleted and the phosphorylnsc kinasc 1~s become inactive. Comparison between these tlT--o procedures in a 2O-fold t7iTuted 30.pellet suspension shops indeed that x discrepancy exists: a rate of 0.16 unit per ml was found with externally added phos-plror~~lasc a as rompared with 0.80 unit per ml cm phosphorylase n included in the complex.
Each of these measurements was made 10 min after flash actirat,ion had been initiat'ed, and when the pho~l)hntase was maximally inhibited; the rates were corrected for dilution of the specific radioactivity and normalized to the same conccntrntioii~ of phosphorylase a since in neither case was the substrate saturating.
One sees, therefore, that el.en in this dissociated system, phosphatase activity was still approximately j-fold greater when measured on endogenous sllbst,rate than when crogcnous substrate was used. When similar measurements m3~e carried out on t,he undiluted 30pellet suspension, a 16.foltl difference in activity was observed (0.35 unit per ml and 6.2 units per ml for the exogenous and endogenous substrates, respectively).
Again, Talues were corrected and normalized as above. It should be kept in mind, ho\vever, that even untlcr conditions such as these, a drastic difference is obser~-ed between inhibition of phosphorylase phosphntnsc in the intnc<t complex as opposed to the solubilized systern. It can be concluded, then, that t.liis effect is a result of phosphoryl:~sc l~hosphat:~se being included in the complex. .5dditional evidence that phosphorylase phosphntase is being liberated upon dilution was obtained as follows. Phosphatase nctil-it)-in the undilutctl 30-pellet suspension (no nucleotide added) was 1.5 units per ml as compared t,o 0.76 unit per ml after 20.fold tlilution.
(a) If phosphor)-lase kinase and phosphatase were both active at the same time, a wasteful recycling of l)liosphor3-lase n and b would occur, and of more importance, (b) the two enzymes would be acting in ol)pnsition to each other, tlrlls interfering with the phosphorylase b to a conversion.
'This :mtagOnisrn decreases the efficiency of the process. Measurements of pliosplior~lasc phosphatase activity in muscle indicates that the activity of this enzyme does not change during contraction (3). The changes in the concentrations of various cffectors, c.g. MI1 and glucose-6-l', have been postulated as important factors in regulating phosphorylnse phosphatase (4-7).
IMcrminatinn of the .UIP concentration in the intact tissue indicates, however, that this level is approximately 100 times higher than the Ki (i), therefore resulting in complete inhibition of phosphatase. Furthermore, the concentration of AMP does not fluctuate in a manner that would allow n physiologically meaningful regulation of the enzyme (15).
In this work with the protein-glycogen complex it was demonstrated that phosphorylnse phosphatase was reversibly inhibited in s?-nchrnny with phosphnrylase kinase activation.
The factor which links phoaphntase inhibition with kinase activation is calcium.
The coordination of these two processes by this metal ion and the observation that it OCCUR3 at the same concentration that triggers muscle contraction indicates its physiological significance.
The mechanism of phosphorylase phosphntase inhibition is not attributable to nucleotitles since there was a striking absence of any AMP effect on the l~hosphat:isc reaction in the protcinglycogen complex, which was further confirmed with use of a phosphopeptide as substrate.
-\lthough a stable covalently modified enzyme ~-ns not d&e&d in the 30.pellet fraction or in purified syatema,l :I l)ossibility exists that covalent. modification is still involved.
Other components of t,he system might interact wit11 a modified phosphat:ise, or these components might themselves require modification in order to inhibit phosphntase; both of these cases might require high concentrations of proteins in order to be observed and thus would not be seen after high tlilution.
Further evidence for the importance of a lligh protein concentration is the observation that eyen when glycogcn is Ii)-drolyzetl, the other components of the 30.pellet fra(%ion protcctj pllo~phorylase phospliatose from IXP inliil)it~ion.
-1lthough added l~l~o~phorylast LC apl~ears not to equilibrate with intrrnal phosphor\-last a, tlicre is still :I coml)lctc loss in the sensitivity of this added substrate to .\MP.
A possible explanation of this effect migllt be tlrat the interaction of added l~hosl~horylnse a with some c~oniponent~s in the complex changes the conformation of this substrate so that it is no longer effected by .0IP.
This would al>o supljort the IlypotlleAs tllat prntcinprotein interactions I)articillate in the over-all phosphorylase system control.