Experimental Depletion of Creatine and Phosphocreatine from Skeletal Muscle*

SUMMARY To evaluate the long term effects on skeletal muscle of feeding a competitive inhibitor of creatine transport, @-guanidinopropionic acid was fed to rats as 1% of their diet for 6 weeks or longer. Although these rats appeared healthy on casual inspection, the P-creatine concentration in their gastrocnemius muscles decreased from a mean value of 22.5 pmoles per g wet weight to 1.6. Instead of P-creatine, the muscles contained a new phosphorylated guanidino compound in concentrations as high as 30 pmoles per g wet weight.

To evaluate the long term effects on skeletal muscle of feeding a competitive inhibitor of creatine transport, @guanidinopropionic acid was fed to rats as 1% of their diet for 6 weeks or longer.
Although these rats appeared healthy on casual inspection, the P-creatine concentration in their gastrocnemius muscles decreased from a mean value of 22.5 pmoles per g wet weight to 1.6. Instead of P-creatine, the muscles contained a new phosphorylated guanidino compound in concentrations as high as 30 pmoles per g wet weight.
By chromatography on Sephadex QAE-A25 this new compound was indistinguishable from phosphorylated pguanidinopropionate formed in vitro in a reaction mixture containing P-guanidinopropionate, rabbit muscle creatine kinase, and an ATP generating system. Hydrolysis of the new compound liberated /I-guanidinopropionic acid and orthophosphate in a 1: 1 molar ratio. When muscles from rats fed P-guanidinopropionic acid were caused to contract under anoxic conditions, the concentration of phosphorylated P-guanidinopropionate decreased dramatically, raising the possibility that it, like P-creatine, can serve in a system to regenerate ATP.
On the other hand /?-guanidinopropionate is a relatively ineffective substrate for creatine kinase in vitro, and animals fed P-guanidinopropionic acid have subnormal concentrations of glucose 6-phosphate, ATP, and ADP in resting muscle in vivo.
These animals may prove useful in studies of the metabolic adaptations of skeletal muscle.
Creatine and P-creatine are known to have several potentially critical roles in skeletal muscle. With creatine kinase, P-creatine participates in a system that regenerates ATP during contraction (2); P-creatine inhibits P-fructokinase (3) and pyruvate kinase (4) and activates fructose I, 6diphosphatase (5) ; and creatine in the bathing medium of skeletal muscle cells in tissue culture stimulates incorporation of [3H]leucine into myosin heavy chain (6). Despite knowledge of these roles, however, it is not certain * This work was supported in part by Research Grant HLO9132 from the National Institutes of Health. This paper is number 6 in a series: Creatine Metabolism in Skeletal Muscle; the preceding paper in the series is Reference 1.
that high concentrations of creatine and P-creatine in skeletal muscle are essential either for function or for viability.
It is true that low concentrations of these compounds are associated with pathologic changes in the muscles of subjects with muscular dystrophy and other myopathies (7), but in no case has it been possible to determine which came first, the abnormality of creatine metabolism or the muscle pathology.
To evaluate the essentiality of high concentrations of creatine and P-creatine in skeletal muscle, a specific means of chronically depleting muscle of these two compounds is needed. Therefore, the present feeding trials with fl-guanidinopropionic acid were initiated.
@-Guanidinopropionic acid is related to creatine structurally and is known to compet,e with creatine for transport into skeletal muscle of the rat (8). Since creatine is not synthesized in muscle and must be transported in from the outside (9), we reasoned that the concentrations of creatine and P-creatine could be lowered specifically by feeding P-guanidinopropionic acid (8). Indeed by the time we began our feeding trials, Shields (10) had demonstrated that feeding P-guanidinopropionic acid to rats as 1% of their diet would lower creatine concentrations in skeletal muscle. Shields also observed that this diet had only a minimal effect, if any, on ability to perform in an exercise test and did not reduce the growth rate of male rats (10). The diets were fed for 6 weeks or longer before the present studies were performed.
To obtain resting muscle, the rats were anesthetized with pentobarbital, the muscles were exposed surgically using care not to provoke contraction, and biopsies were taken with rongeurs prechilled in liquid nitrogen.
These biopsies were kept solidly frozen as they were weighed, pulverized, and mixed with frozen perchloric acid prior to preparing a neutral extract for measurement of muscle phosphates.
The details of preparation of the muscle extract and of the automated, chromatographic method used to separate and quantitate muscle phosphates have been published (11). Briefly, the extract is applied to a column of Sephadex QAE-A25, and the phosphate compounds are eluted with a mixture of solutions which form a continuous gradient of decreasing sodium sulfate A brief description of methods is given in the text. The elution pattern obtained by monitoring for phosphate is shown in the upper traci?lg for resting muscle of a normal rat and in the middle Iracitlg for resting muscle of a rat fed p-guanidinopropionic acid; the peaks representing glucose G-phosphate, P-creatine, Pi, ADP, and ATP are identified. The elution pattern obtained by monitoring for Sakaguchi positive compounds is shown in the bottorr~ tracircg for resting gastrocnemius muscle of a rat fed p-guanidinopropionic acid; the first peak corresponds to @-guanidinopropionate and the other peak represents a new guanidino phosphate.
Neither of these guanidino peaks was present in the elution patterns of extracts from normal muscles.
As the eluate leaves the column, the organic phosphates are partially hydrolyzed first by treatment with NaOH and subsequently with a HzS04molybdate reagent.
The phosphate content of the cluatc is then continuously monitored using the Fiske-SubbaRow method adapted for the AutoAnalyzer (11, 12). To simultaneously measure monosubstituted guanidino compounds, the cluate from the column was split into two streams after the second hydrolysis step, and one of these streams was continuously monitored by the Sakaguchi reaction adapted for use with the AutoAnalyzer. The methods employed to develop and stabilize the color produced in the Sakaguchi reaction were the same as those described by Van I'ilsum et al. (13) except that caffeine was substituted for thymine in one of the reagent,s. The outputs from monitoring colorimcters were recorded to obtain elution patterns (Fig. I), from which the concentrations of the various compounds were calculated.
Enzymatic measurements of Kcreatine (14) and l'i (15) were made in certain instances, using enzymes and other reagents obtained from Sigma Chemical Company; and the total creatine contents of some of the muscle samples were measured by the method of Rose et al. (16) after ascertaining that fi-guanidinopropionic acid causes no interference in this method. Total water content was determined by drying muscle samples to constant weight at 85".
The ability of P-guanidinopropionate to serve as a substrate for creatine kinase was studied in the coupled spectrophotometric system of Tanzcr and Gilvarg (17) as modified by McLaughlin et al. (18) for kinetic studies. The enzymes and other reagents for this system were purchased from Sigma Chemical Company.

AND DISCUSSION
Although ihe rats fed P-guanidinopropionic acid appeared healthy on casual inspection throughout the feeding trials, the I'-creatine concentrations in their gastrocnemius muscles were  (7) a Resting gastrocnemius muscles from young rats weighing less than 250 g and of both sexes were used for these studies.
* Mean f S.lS.; number of animals in parentheses.
Except for Pi, all of the differences between normal and experimental animals have a p value of 0.02 or less as determined by the Student's t test.
c The amount of Pi could not be assessed from the elution patterns from experimental animals (see Fig. 1). reduced to less than 10% of the normal value ( Fig. 1 and Table  I). This reduction in concentration was apparent from an examination of the elution patterns from the Sephadex QAE-A25 column (Fig. I), and it was confirmed by enzymatic measurements of Kcreatine (Table I). In agreement with the work of Shields (lo), the total content of creatine also was reduced from 38.7 f 1.7 (mean & S.E.) pmoles per g wet weight of muscle for nine normal rats to 9.62 =t 0.92 for seven rats fed P-guanidinopropionic acid. These changes in content of creatine and I-'-creatine are not due to replacement of muscle cells by fatty or connective tissue (lo), and they are not due to changes in water content.
We found the water content to be 76.7 f 0.47, of wet weight of muscle for nine normal rats and 77.6 =t 0.4 for eight rats fed ,&guanidinopropionic acid. Instead of a large P-creatine peak, the elution pattern obtained from muscle extracts of rats fed P-guanidinopropionic acid displayed an exceptionally large peak in the position normally occupied by the I'i peak (Fig. 1). Although the new peak at first was considered to represent l'i (19) subsequent enzymatic measurements revealed normal concentrations of Pi in skeletal muscle of rats fed /3-guanidinopropionic acid (Table I). Thus, the existence of a new organic phosphate was apparent.
From the following observations we conclude that the new organic phosphate is phosphorylated @-guanidinopropionate. (a) Hydrolysis of that fraction of the eluate which contained the new compound liberated l'i and a Sakaguchi positive compound' in a 1: 1 molar ratio (Fig. 1). (b) Acid hydrolysis of the muscle extract prior to Sephadex QAE-A25 chromatography liberated a Sakaguchi positive compound that chromatographed as P-guanidinopropionate.
The position of the P-guanidinopropionate peak in the chromatographic tracing is shown in Fig. 1. In this system fi-guanidinopropionate is easily distinguished from the new compound and from arginine because it emerges from the column much sooner than the new compound and slightly later than arginine.
fl-Guanidinopropionate and guanidinoacetate are not separated by the Sephadex QAE-A25 column.
(c) By paper chromatography using two different solvent systems, the major Sakaguchi positive compound in muscle extracts from rats fed P-guanidinopropionic acid behaved identically to authentic 1 The quantity of Sakaguchi positive compound was estimated from a standard curve prepared from authentic p-guanidinopropionic acid. The reaction mixture consisted of ATP, 3 mM; MgSOa, 6 mM; P-enolpyruvate, 1 mg per ml; DPNH, 0.1 mg per ml; glytine, 100 mM, pH 9; rabbit muscle lactate dehydrogenase, 0.33 mg per ml; rabbit muscle pyruvate kinase, 0.5 mg per ml; and creatine or @-guanidinopropionate to achieve the concentrations shown in the figure; 1 mg of rabbit muscle creatine kinase per ml was added at-4 min.
This mixture was incubated at 25" under room air, and the change in optical density at 340 nm was recorded as a function of time.
&guanidinopropionic acid (20) and was easily distinguished from arginine and guanidinoacetic acid (Table  II) muscle creatine kinase under the conditions of the assay. This result is in agreement with the knowledge that phosphorylated guanidinoacetate is a poor substrate for this enzyme (21). To determine whether phosphorylated P-guanidinopropionate is labile under conditions known to deplete P-creatine in vivo, gastrocnemius muscles were caused to twitch under anoxic conditions (Fig. 3). As expected for normal muscle (Fig. 3), the P-creatine concentration fell to very low levels and there were reciprocal increases in glucose &phosphate and I'i. The mean total loss of P-creatine and ATP from normal muscle was 15.9 pmoles per g. Under the same conditions the total loss of phosphorylated fl-guanidinopropionate and ATP from muscles of rats fed P-guanidinopropionic acid was 12.0 pmoles per g (Fig. 3). Both the initial and final concentrations of phosphorylated /% guanidinopropionate were higher than the corresponding values for P-creatine.
Again there were reciprocal increases in glucose 6-phosphate and Pi. These findings clearly demonstrate that phosphorylated P-guanidinopropionate is labile under the conditions produced by anoxic contraction of skeletal muscle, and they raise the possibility that this compound serves in a system to regenerate ATP in a manner similar to P-creatine.
The ability of phosphorylated @-guanidinopropionate to function in such a system could explain apparent good health despite severe depletion of Pcreatine from skeletal muscle of rats fed P-guanidinopropionic acid. Additional work is needed to prove that phosphorylated P-guanidinopropionate in fact does substitute for P-creatine.
Studies are in progress to evaluate the metabolic adaptations consequent to substituting /3-guanidinopropionate for creatine in skeletal muscle. That metabolic adaptations occur is evident from the reduced concentrations of glucose 6-phosphate, ADP, and ATP in resting skeletal muscle of rats fed P-guanidinopropionic acid ( Table I). Elucidation of these adaptations should add to our understanding of the roles of creatine and I'-creatine in the regulation of metabolism in skeletal muscle in Go.