Biochem. J. (1970) 118, 887-893Printed in Great Britain

887

The Meyerhof Quotient and the Synthesis of Glycogen from Lactate inFrog and Rabbit Muscle

A REINVESTIGATION

By J. R. BENDALL AND A. A. TAYLORAgricultural Research Council Meat Research Institute, Langford, Bristol BS18 7D Y, U.K.

(Received 25 February 1970)

1. The conversion of lactate into glycogen was demonstrated in frog sartoriusmuscle in oxygen. The rates and amounts are highest when lactate is added to thebathingmedium and are dependent on lactate and CO2 concentration, as well as pH.The glycogen content of a resting muscle can be doubled in 4h at 24°C. 2. Sartoriusmuscle, recovering aerobically in liquid paraffin from a period of anoxia, convertspreformed lactate intp glycogen at a lower rate and in smaller amounts than whenlactate is added in an aqueous medium. The lower rates are similar to those Meyerhoffound under the same conditions, after correction for temperature; they can beattributed partly to low muscle pH and partly to the limited amounts of lactatepresent. 3. Rabbit psoas muscle also shows the ability to convert added lactate intoglycogen under aerobic conditions. The rates are low and similar to those in frogsartorius muscle recovering from anoxia. 4. The present experiments yield a

Meyerhof quotient of 6.2, compared with Meyerhof's value of 4-5. However, thesevalues are not significantly different from one another. 5. It is suggested that theglycogen coefficient, i.e. mol of glycogen formed/mol of lactate disappearing, is amore reliable way ofassessing the resynthetic mechanism than the original quotient,i.e. mol of lactate disappearing/mol of lactate oxidized. The found coefficient is0.419±0.024.

Meyerhof (1920a, 1922) was the first to discover afeature ofoxidative recovery in frog muscle in whichsome ofthe lactate formed during fatigue disappearswhen the muscle is allowed to recover in oxygen,but only about 20-25% of this lactate can be ac-counted for by oxidation to CO2 and water.Meyerhof (1920a, 1922) expressed his results in theform of the quotient (total lactate disappearing)/(lactate oxidized), the numerator of which wasbased on direct measurement of lactate and thedenominator on the 02 uptake divided by 3.Altogether Meyerhof (1920a, 1922) recorded 21separate experiments on muscles fatigued in variousways; a re-plot of these data in their raw form yieldsa quotient of 4.1±0.383 (s.E.M. for 19 degrees offreedom), i.e. only about one-quarter of the missinglactate could have been oxidized. Two laterexperiments ofMeyerhof & Lohmann (1926) yieldedquotients of 4 and 6. Thus there can be no doubtabout the general validity of the quotient.Meyerhof (1920a, 1922) concluded from his earlier

experiments that the lactate unaccounted for byoxidation had been converted into carbohydrate,mostly in the form of glycogen. In that case themean value of the quotient given above shows that0.38±0.13mol of glucose units of glycogen or of

other carbohydrate should have been resynthesized/mol of lactate disappearing. To confirm this hypo-thetical formation of carbohydrate, Meyerhof(1920a, 1922) seems to have carried out only twodirect experiments, in both of which the gains inglycogen did, in fact, account for most ofthe missinglactate. However, the experiments were of verylong duration, and it can be questioned whether thiswas a sufficient basis on which to build the far-reaching and oft-quoted hypothesis that finallyemerged. The only other confirmatory evidenceseems to be that of Foster & Moyle (1921) andMeyerhof, Lohmann & Meier (1925). The latterauthors perfused the hind limbs of frogs with 0.12%lactate and reported significant gains in the glycogencontent of the muscles. Eggleton & Evans (1930)repeated these latter experiments but could findonly small and erratic gains in glycogen which werenot statistically significantly different from zero.

There seem to have been very few further attemptsto repeat Meyerhof's experiments until quite recent-ly, e.g. Bar & Blanchaer (1965) and Gourley & Suh(1969). The latter authors showed some glycogenformation from added lactate in isolated sartoriusmuscles, but the lactate concentrations theyemployed (3mM) were far below that to be expected

J. R. BENDALL AND A. A. TAYLOR

in a fatigued muscle. There is thus considerablejustification prima facie for the severe criticism ofthe entire hypothesis by Krebs & Woodford (1965),who objected on the grounds of the thermodynamic'barrier' to glycogen formation which would existunless the following enzymes were present: L-malate-NADP oxidoreductase (decarboxylating)(EC 1.1.1.40), or alternatively pyruvate carboxylase(EC 6.4.1.1), phosphoenolpyruvate carboxykinase(EC 4.1.1.32) and fructose diphosphatase (EC3.1.3.11). They were able to demonstrate only thepresence of the last of these in 'white' muscle.However, Opie & Newsholme (1967) later discoveredsignificant quantities of the first and third of theseenzymes, so the distinct possibility now exists thatthey are involved in the first steps of the conversionof lactate into glycogen. In view of this it seemedtimely to repeat Meyerhof's experiments, with themore refined techniques of assay now available.The results demonstrate that glycogen can beformed in large amounts from lactate, added toisolated frog sartorius muscles.

MATERIALS AND METHODS

Preparation of muscle samples. The sartorius musclesofEnglish frogs (Rana temporaria) were carefully dissectedout, cut free from their origin and tendon of insertionand placed in Warburg vessels with 1 ml of medium (seebelow) until required.

In some experiments the muscles were dissected outimmediately after the death ofthe animals by decapitationand pithing; in others the animals were pinned out afterdeath on boards, placed in large bottles with moist N2,sealed and stored in a room at 100C for 16-20h to allowanaerobic accumulation of lactate. The muscles were thendissected out as above.

Ringer solution and other bathing fluids. When Ringersolution was used, its composition was as follows (Hill,1965, p. 244): NaCl, 115.5mM; KCI, 2.5mM; CaCl2,1.8mm; phosphate buffer, 3mM, at pH7.0 (sodium salt).In several experiments lactate was added to it in variousconcentrations as noted in the text. It was prepared frompure lactic acid (Sigma Chemical Co., St Louis, Mo.,U.S.A.), neutralized with NaOH, and the lactate contentwas assayed enzymically (see below).

In some experiments, designed to reproduce the condi-tions of Meyerhof's (1920b) experiments, the muscles weresuspended in neutral paraffin, density 0.83-0.87 (BDHChemicals Ltd., Poole, Dorset, U.K.), previously saturatedwith water (Meyerhof, 1920b).

Measurement of oxygen uptake. The 02 uptake wasmeasured in a standard Warburg apparatus at 24°C,usually with 25% KOH in the central well ofthe vessels toabsorb the C02 formed.The sartorius muscles of ten frogs were used in each

experiment and after being weighed were placed in pairsin Warburg vessels containing 1 ml of medium. They werethen equilibrated with pure 02 for 15min at the bathtemperature of 240C. The arrangement of the musclepairs, in order of killing the frogs, was as follows (R, right

sartorius; L, left sartorius): vessel no. 1, muscle no. 1 L,6R; vessel no. 2, muscle no. 2R, 7L; vessel no. 3, muscleno. 3L, 8R; vessel no. 4, muscle no. 4R, 9L; vessel no. 5,muscle no. 5L, 1OR; vessel no. 6, muscle no. IR, 6L;vessel no. 7, muscle no. 2L, 7R; vessel no. 8, muscle no.3R, 8L; vessel no. 9, muscle no. 4L, 9R; vessel no. 10,muscle no. 5R, 10L. Immediately after equilibration with02, the contents of vessels 1-5 were combined and addedto 6ml of 50% KOH solution for assay of initial glycogenand lactate content. Vessels 6-10 were allowed to take up02 for various times and then their contents combined andplaced in KOH as above.

In some experiments, where it was desired to allow C02to accumulate during the experiment, KOH solution wasomitted from the central wells ofthe vessels. The approxi-mate 02 uptake was then measured from the observedpressure changes by Warburg's direct method, assuminga respiratory quotient of 1.0 (Umbreit, Burris & Stauffer,1947) and correcting for the pH of the bathing solutions.The initial pH varied between 6.9 and 7, and the final pHbetween 7.3 and 7.54. The latter values are probably toohigh because of 18o of C02 during measurement.Assay of glycogen. The muscle samples, after digestion

in hot 25% KOH, were prepared for determination ofglycogen by the method of Good, Kramer & Somogyi(1933). Glycogen was hydrolysed to glucose in 1.2M-HCIfor 1 hat 100°C. The glucose was measured, after oxidationby alkaline ferricyanide, by back-titration with cericsulphate; Xylene Cyanol FF (BDH Chemicals Ltd.) wasused as an internal redox indicator (modified method ofCole, 1944). The method is capable of measuring as littleas 0.01 ,umol of glucose. For this reason it was preferredto the more usual anthrone method. Experiments inwhich glycogen was added to the samples showed thatthe recovery was 99-101% over the range of glycogenconcentrations actually encountered in practice (15-17,umol/g of muscle).Assay of lactate. Lactate was assayed enzymically by

using lactate dehydrogenase and NAD+ at alkaline pHand measuring the NADH formed spectrophotometricallyat 340nm (Hohorst, 1963). By this method lactate can beaccurately assayed in the KOH-ethanol supernatantsleft over from glycogen assays, after neutralization withHCl04, decantation from the KCl04 precipitate andappropriate dilution (R. K. Scopes, personal communica-tion). The recovery of lactate from test samples was100±1%. This simple procedure obviates the need todouble the number of muscle samples for the two separateassays, and thereby decreases very considerably the vari-ability in the results.

Assay of hexose 6-phosphate. This was assayed enzymic-ally by the method of Hohorst (1963). Because the assayhad to be made on neutralized HC104 extracts of sartoriusmuscles the preparations used for the 02 uptake andglycogen and lactate assays were not suitable. Separategroups of frogs were therefore used for the assay ofhexose 6-phosphate.

RESULTS

Experiments with sartorius Muscles from freshlykilled frogs. The experiments of Gourley & Suh(1969) showed that some glycogen formation from

1970888

Table 1. Experiments with sartorius muscles offreshly killedfrogs at 24°C, with and without addition of lactateto the Ringer solution

The parameters are given in terms of ,umol/g of muscle, except where noted. The initial hexose 6-phosphatecontent was approx. 3.5 tmol/g. The mean weight of a muscle pair was 0.15g (in 1 ml of Ringer solution). Fivepairs were combined for initial assays and five for final assays. The initial pH varied between 6.95 and 7.05.

Expt. no.Duration (h)[Lactate] added in Ringer (mm)Initial [lactate] in Ringer+muscle(mm)

Lactate disappearing (A) (,umol/g)Initial glycogen (glucose units)Glycogen synthesized (B)02 uptake (C)Calc. lactate disappearing(D= C/3+2B)CO2 absorbed into KOH or not

13.5003.0

7.016.8-2.933.75.5

2

24.003.2

17.826.0

-23.7

34.725.07.4

23.224.8+1.343.117.0

4 5 6 74.00 4.25 4.2538.0 18.0 38.039.0 18.2 39.0

26.029.8+8.156.234.9

42.524.6

+13.649.643.7

60.030.2

+19.6

Yes Yes Yes Yes No No No

lactate occurred in frog muscle, even at the verylow concentration of lactate they employed (3mm).Our own preliminary experiments confirmed thisand suggested that much larger amounts ofglycogenwould be formed at higher lactate concentrations.In the following experiments therefore concentra-tions of lactate from 3 to 50mM were added to theRinger solution bathing the muscles, respiratoryCO2 being allowed to accumulate in some cases,whereas in others it was absorbed into potassiumhydroxide.The marked effect of lactate concentration on the

formation of glycogen and on lactate utilization isseen clearly from the results in Table 1 for sartoriusmuscles removed immediately after death. When nolactate was added to the Ringer solution, as inExpts. 1 and 2, and the initial lactate concentrationwas about 3mM, glycogen was actually lost from themuscle despite oxidation of some of the preformedlactate. In such a case the loss of glycogen anddisappearance of lactate in relation to the 02

taken up indicated that both these sources of C3compounds were being oxidized viathe tricarboxylicacid cycle. However, when the lactate concentra-tion was increased above 7mM, larger amounts ofglycogen were synthesized; at 39mM-lactate 8.1,umol (1.45mg) of glycogen (glucose units) was

synthesized/g of fresh muscle in 4h in the absenceof respiratory C02, whereas in its presence theamount increased to 15-19.6,umol (2.7-3.5mg)/g.The latter value represents a doubling of the initialglycogen content at the expense of added lactate.C02 had a marked effect on the rate of glycogen

resynthesis. A possible explanation is that it isneeded for the operation of the 'malic' enzyme[malate-NADP oxidoreductase (decarboxylating)],which might be the first step in the resyntheticchain from pyruvate to glycogen. The amount of

CO2 actually available will depend on the pH of thesolution bathing the muscle. In Expt. 6 in Table 1the initial pH was 6.97 and the final pH7.54. As-suming both the rise inpH and the liberation of C02were linear with time, the total C02 liberated half-way through the experiment was about 35,umol/gof muscle and the pH was 7.25. Under the actualconditions of 0.17g of muscle in 1ml of Ringersolution with a gas head-space of 19ml and a

temperature of 240C, the calculated maximalconcentration of C02 + HCO3- in the Ringersolution was 0.2mM. It would in fact be lower thanthis, because the measured final pH is probably toohigh (see the Materials and Methods section).To decide to what degree the spontaneous

accumulation of C02 was optimum, two identicalexperiments were carried out, in which sodiumhydrogen carbonate was added to the Ringersolution to a final concentration of 11.9mM, and an

atmosphere of 02+C02 (95:5) was used instead ofpure 02. Lactate was also added to a concentrationof 42mM; the initial pH was 7.10 and the final pHabout 7.5. The total concentration of C02+HCO3-was initially 13.6mM (about 68 times that inExpt. 6 in Table 1). The amounts of glycogenfoimed/g in 4jhwere 28.0 and 24.6,umol of glucoseunits. When these values are compared with thatof 19.6,umol/g in Expt. 6 in Table 1 it is clear thateven the low concentration of C02 present duringthe latter experiment (0.2mM) was at least 75%optimum.The combined results of these lactate addition

experiments cover a wide range of glycogen syn-

thesis, and therefore serve to establish the generalrelation between glycogen synthesized and totallactate disappearing. It was found that glycogensynthesis was linear with time, providing the lactateconcentration in the bathing fluid did not fall

4.2541.542.5

50.031.1

+15.0

Vol. 118 M.EYERHOF QUOTIENT IN MUSCLE 889

J. R. BENDALL AND A. A. TAYLOR

+ 7

+ 6

0++4

_+30

+2

-+1

0-

o-20

-ALactatE

Fig. 1. Plot oftheralAt) against the ratein freshly excised frBathing solution wahad been added in vEThe temperature w.glycogen formed andisappearing estimal*, Rabbit experime

appreciably durinreason the obser

Fig. 1 as rates, togglycogen synthesithat the measur

utilized to oxidizethis amount oflagives a measureo

cogen. It is seen

values agree rea

observed paramel

shown in Fig. 1 E

regression line forThe regression

glycogen synthesi,in freshly excised

AGlycogen/At =

r = 0.99 for n = 9

±0.024.The additive

metabolic rate, i.Evia pyruvate and

supply the energbroken down slovLawrie, 1962).

Experiments oaoxygen from 16-described above wi

glycogen synthesi:in frog muscle ur

have seen, thean;

The actual conditions obtaining in Meyerhof's(1920a, 1922) original experiments were, however,0 very far removed from this optimum state, because

the muscles had usually been tetanized for upwardsof 15min at 140C so that they could accumulate asmuch lactate as possible. The result of this ratherfierce treatment was to lower the pH to about 6.5

0 and to decrease the creatine phosphate and ATPreserves ofthe muscle almost to zero. By suspending

o the muscles in an atmosphere ofmoist 02 Meyerhof(1920a) also limited the amount of lactate availableto the intrinsic value of about 30,umol/g, and the

I Ii possible resynthesis of glycogen to somewhat less5 10 15 20 than 12,umol of glucose units/g. In practice, he

etotal/At (,umol of lactate/g per h) never obtained a resynthesis of more than 8,umolof glucose/g [see Meyerhof (1920a), p. 313, for a

teof lactate disappearance(ALactate/ direct estimation, and Tabelle III, expt. 3, p. 294,f glycogen formatlon (eGlycogen/dt) that be derived indirectly].

s Hll's frog Rilnger towh0chlactateWe have preferred to Meyerhof's (1920a) pro-

arious amounts (see Table 1 and text). longed and vigorous electrical stimulation theas 24°C. o, Direct measurement of gentler procedure of leaving the muscles intact inid lactate disappearing. A, Lactate the decapitated animals for 16-20h under anaerobic,ted from 2AGlycogen/At-(A02/At)/3 conditions at 100C; by this means nearly as muchnts from Table 3. lactate accumulates, due to the resting metabolism,

as by Meyerhof's (1920a) method. We also usedMeyerhof's (1920b) technique of placing the

ng the experiment. For this muscles in pairs in 1 ml of paraffin, instead of sus-

ved parameters are plotted in pending them in moist 02, during the 02 uptakerether with the calculated rates of experiments. This avoids drying out of the muscless, based on the assumptions: (a) and provides a medium with a very high 02solu-ed 02 uptake was exclusively bility. In spite of the long holding time, the9

lactate; (b) that subtraction Of bacterial counts of muscles under these conditionsotate from the total disappearing are less than 106/g at the end of the experiment.f the lactate converted into gly- The results are summarized in Table 2, whichthat the observed and calculated shows that there is some glycogen resynthesis insonably well. Values for the every case, although the amounts are much smallerters in rabbit psoas muscle are than might have been expected from Table 1. Asnd are seen to fall close to the we implied above, this is probably attributablefrog sartorius muscle. partly to the limited amount of lactate available,line relating observed rates of but mainly to the low initial pH values (6.45-6.76,

s to rates of lactate disappearance compared with about 7.0 in Table 1) and the lowerfrog muscle is given by: creatine phosphate and ATP contents.

-0.419(ALactatetotail/At) - 1.3 (1) To confirm the effect of pH, we carried out twoexperiments of the same type as those in Table 1,

S.E.M. of regression coefficient= but on muscles allowed to accumulate lactateanaerobically for 16h or so. A 1 ml volume of

constant represents the resting Ringer solution was used as bathing medium anda. the rate of glycogen oxidation lactate was added in one case to a final concentra-the tricarboxylic acid cycle to tion of 21 mm and in the other to 40mM.C02was

gy for resynthesis of the ATP allowed to accumulate in both cases. The rates ofvly in resting muscle (Bendall & resynthesis of glycogen were respectively 2.55 and

2.84tLmol/g per h and the pH was initially aboutIsartorius muscles recovering in 6.5; these values compare with 3.2 and 4.6,umol/-20h anoxia. The experiments g per h for the experiments in Table 1 (Expts.5 and 6ere carried out to discover whether respectively), which had an initial pH of about 7.0s from added lactate was possible but were otherwise under the same conditions.ider optimum conditions; as we In Table 2, it is noteworthy that Expts.10 and11zwer is clearly in the affirmative. show a considerable discrepancy between the

890 1970

MEYERHOF QUOTIENT IN MUSCLETable 2. Experiments with sartorius muscles recovering in oxygenfrom 16-20h anoxia

The temperatiore was 240C. The suspension medium was 1 ml of paraffin. The mean weight of a muscle pairwas 0.16g. The initial hexose 6-phosphate content was 3-6,umol/g. All parameters are given in terms of tmol/gof muscle.

Expt. no.Duration (h)Initial pHInitial lactateLactate disappearing (A)Initial glycogenGlycogen synthesized (B)02 uptake (C)Calc. lactate disappearing(D = C/3+2B)C02 absorbed into KOH

8 9 10 114.40 4.17 5.00 2.006.45 6.75 6.76 6.58

26.8 20.0 19.3 23.56.2 9.7 14.3 5.7

15.2 20.8 28.9 26.71.3 1.7 5.3 3.1

46.0 13.7-- 25.9 10.9

No No Yes Yes

Table 3. Experiments with strips ofpsoas major muscle offreshly killed and immobilized rabbits (see the text)Lactate was added to Ringer solution, as shown. The mean weight of a strip was 0.2g. Four strips were

combined for initial assays and four for final assays. Parameters are in ,umol/g, unless otherwise stated.

Expt. no.

Duration (h)Initial pH of Ringer plus muscle[Lactate] added in Ringer (mm)Lactate disappearing (A)Initial glycogen (glucose units)Glycogen synthesized (B)

02 uptake (C)Calc. lactate disappearing(D = C/3+2B)C02 absorbed into KOH

12(normal)

4.177.02

43.318.062.43.0

(= 0.72/h)34.817.6

(= 4.2/h)Yes

13

6.007.10

47.0

8.95.7

(= 0.95/h)37.323.8

(= 4.0/h)Yes

14

5.007.09

46.015.02.14.3

(= 0.86/h)33.019.6

(= 3.9/h)No

measured disappearance of lactate and that neededto account for the 02 uptake plus the synthesis ofglycogen. It is possible that this discrepancy wasmade up by utilization of hexose 6-phosphate as anextra source of lactate or pyruvate. Under theconditions of these experiments, the initial hexose6-phosphate contents lie between 3 and 6,umol/g,which would yield double these contents of lactateequivalents. This would satisfactorily explain thediscrepancy, which only plays a major role whenthe amount oflactate available is limited in thiswayby the particular experimental design.

Glycogen synthesi8 from lactate in rabbit psoasmu8cle. For these experiments the rabbits wereparalysed with curare for 10min before death toensure a high initial content of ATP and creatinephosphate. Two of them had also been injectedwith 2ml of 0.1% adrenaline lih before death todecrease the glycogen content (Bendall & Lawrie,1962) whereas the third was a normal well-fedanimal. Strips were prepared from the psoas majormuscle (each about 5cm long and weighing about0.2g) and placed in 1 ml of Krebs-Ringer phosphatesolution, with the addition oflactate to a concentra-tion of 43-47mm. Four strips were combined for

determination of the initial parameters, and fourwere allowed to take up 02 for various times andthencombined and assayed in the same way (see theMaterials and Methods section).As seen from Table 3, all the muscles synthesized

about the same rather low amount ofglycogen fromlactate, although in Expt. 12 the initial glycogencontent was 7 times that in Expt. 13 and 30 timesthat in Expt. 14. All utilized lactate at a rate ofabout 4,umol/g per h (accepting the calculatedvalues as a basis). The amounts of glycogen syn-thesized are about halfthose shown by frog sartoriusmuscles under identical conditions (cf. Expt. 4 inTable 1) andnearly the same as those ofthe sartoriusmuscles recovering from a period of anoxia (Expts.10 and II in Table 2). When the results are plottedas in Fig. 1, as glycogen synthesized against totallactate disappearing, the points (as solid circles)fall quite close to the regression line for frogsartorius muscles.

DISCUSSION

Thepresent results appear to vindicateMeyerhof'soriginal hypothesis in its two most important

Vol. 118 891

J. R. BENDALL AND A. A. TAYLOR

aspects: first, that the total amount of lactate dis-appearing from frog muscle during aerobic recoveryfrom a tetanus or a period of anoxia is 5-6 times theamount that is oxidized; secondly, that the missinglactate reappears mostly in the form of glycogen.It is the latter point that has been most in disputebecause of its great theoretical interest (Eggleton &Evans, 1930; Krebs & Woodford, 1965).

In discussing the synthesis of glycogen fromlactate in muscle, we must distinguish clearlybetween what is possible under optimum conditionsof high pH, lactate and CO2 concentrations andwhat occurs during recovery from actual anoxia.Optimum or near-optimum conditions were clearlypresent at the higher lactate and C02 concentra-tions shown in Table 1 (Expts. 5, 6 and 7) and in twoexperiments mentioned in the text where an atmo-sphere of 02+C02 (95:5) was substituted forpure 02. These results, shown as the upper pointsin Fig. 1, demonstrate that between 13.6 and28.0,umol of glucose units of glycogen/g can beformed from lactate added to a respiring sartoriusmuscle during an experimental period of about4ih giving in the best cases a final glycogen contentof 1.1% ofthe muscle weight. The latter is certainlya high value for frog muscle (Meyerhof, 1920a, 1922;Eggleton & Evans, 1930) and is about the same asthe initial glycogen content of many restingmammalian muscles, e.g. rabbit psoas muscle at1.0-1.3% and ox longissimus dorsi muscle at 1.5-2.0% (J. R. Bendall, unpublished work). Thus therecan be no doubt that glycogen can be synthesizedfrom lactate by a respiring frog sartorius muscle inamounts that are much too large to be accountedfor by the thermodynamically unfavourable routeof mere reversal of the normal glycolytic pathway(Krebs & Woodford, 1965).The question how important the aerobic resyn-

thesis of glycogen may be in an isolated muscle,recovering from a tetanus or from anoxia, is notquite as simple to answer as the above remarksmight indicate, although it is clear from Table 2that such resynthesis does occur to a limited extent,but at a rather low rate. For example, the resyn-thesis observed in such isolated muscles was only5.3,umol in 5h at 240C, compared with 28,pmol in4h under the optimum conditions discussed above.Such amounts are of about the same magnitude aswould be expected from Meyerhof's (1920a, 1922)results under similar conditions, after correctingfor the temperature difference. As pointed out inthe Results section, the lower rates are probablymainly due to the lowerpH (approx. 6.5) obtainingin muscles recovering from a period of anoxia.Even a white muscle of the rabbit, the psoas

major, appears to be capable of resynthesizingglycogen from lactate to some degree, although theamounts are small even under optimum conditions

of high pH and high lactate and C02 concentrations(see Table 3). 0

Having demonstrated that glycogen can beresynthesized from lactate in respiring frog andrabbit white muscle, the next question that arisesis the stoicheiometric relation between the twoparameters. As we have shown, 0.419mol of gly-cogen (glucose units) is synthesized/mol of lactatedisappearing in frog muscle (see Fig. 1 and eqn. 1);this we have termed the glycogen coefficient. Itcan be readily converted into the more conventionalMeyerhof quotient (QM), since it follows from itthat, per 1 mol of lactate disappearing, 0.838mol isformed into glycogen and 1-0.838 = 0.162mol isoxidized to provide the necessary energy. ThereforeQM= 1/0.162 = 6.2. From the standard error of theglycogen coefficient (±0.024), however, the con-fidence limits for the quotient at the 5% level ofprobability are set at 3.6 and 20.0, and even at the25% level they are decreased only to 4.5 and 9.8.Meyerhof's own results show that where the

lactate disappearing and the lactate oxidized weredirectly measured, the mean value of QM was 4.1(Tabellen II, III and V in Meyerhof, 1920a). Thestandard error for the 21 values available is ±0.383,giving confidence limits of 4.9 and 3.3 at the 5°%level of probability. The narrower limits set to QMin this case are due to the fact that the parameterswere directly measured, whereas to convert ourvalues into these terms it is necessary first tosubtract a large value (0.838) from 1.0 and then totake the reciprocal, which automatically decreasesconfidence in the derived quotient. The result isthat our QM value does not differ significantly fromMeyerhof's (1920a, 1922) despite the apparentlylarge difference between them. It is not worthwhiletaking this argument further, since the QM is of nodirect use in determining the nature of the reactionsthat underlie the process.

All the enzymes necessary for glycogen synthesisfrom lactate have been demonstrated to be presentin frog muscle (Opie & Newsholme, 1967) includingthe so-called 'malic' enzyme, which actuallyconsists of L-malate-NADP oxidoreductase (de-carboxylating) coupled with L-malate-NAD oxido-reductase (EC 1.1.1.37). These steps need C02 tooperate in the required direction, but not ATP. Analternative pathway might be via pyruvate carbo-xylase, which requires both C02 and ATP, but it isextremely doubtful whether this enzyme is presentin muscle in significant quantities (Opie &Newsholme, 1967). The next enzyme required,after these initial steps, is phosphoenolpyruvatecarboxykinase, which the latter authors also showedto be present. This requires GTP, which can be re-synthesized from ATP via the ubiquitous nucleosidediphosphate kinase (EC 2.7.4.6). Lastly, but muchlater in the chain, there is a requirement for fructose

892 1970

Vol. 118 MEYERHOF QUOTIENT IN MUSCLE 893

diphosphatase which Krebs & Woodford (1965)demonstrated in muscle.

If the 'malic' enzyme operates, the overall reac-tion from lactate to glycogen can be formulated asfollows:

2 lactate' + 5 ATP4- + (glucose) +H205 ADP3-+5 p12-+3H++(glucose)n+i (2)

Calc. AFO = -5.4kcal/mol of lactate

As a result of the production of ADP3-, the mito-chondrial electron-transport chain would operate,with either pyruvate or lactate as fuel, to produce17mol of ATP from ADP on oxidation of 1 mol toCO2 and water (assuming that the x-glycerophos-phate shunt is operating). Thus 2.5mol of ATP isused up/mol of lactate converted into glycogen(eqn. 2) and 17mol is produced/mol of lactateoxidized: or for every 1 mol of lactate disappearing,2.5/17 = 0.147mol is oxidized and 0.853mol isconverted into glycogen. Thus the theoreticalglycogen coefficient as defined by eqn. (1) is 0.427against the found value of 0.419. From the givenstandard error of the latter (0.024), the confidencelimits at the 5% level of probability are 0.362 and0.476, so that the difference between the predictedand found values ofthe coefficient is not statisticallysignificant. It is also noteworthy that Meyerhof's(1920a, 1922) meanQM value of 4.1 yields a glycogencoefficient of0.38, but because ofthe wide confidencelimits of L0.05 this value does not differ significantlyfrom either ours or the theoretical value.The alternative route of resynthesis via pyruvate

carboxylase would require 1 more mol of ATP/mol of lactate utilized and would thus yield aglycogen coefficient of 0.397. This again does notdiffer significantly from the found value of 0.419,although a line drawn with this slope fails to passthrough the majority of the experimental points inFig. 1. For these reasons, together with Opie &Newsholme's (1967) failure to find pyruvate carbo-xylase in 'white' muscle, we consider the 'malic'enzyme pathway to be the more probable route ofresynthesis. This would also explain the findingthat CO2 concentration plays an important role inthe overall reaction, since L-malate-NADP oxido-reductase (decarboxylating) has a rather lowaffinity (high Ki) for C02 (Mahler & Cordes, 1966).

Conversion of lactate into glycogen may play apart in contraction experiments with isolatedmuscles, such as those reported so extensively byHill (1965, pp. 183, 200). In such experimentslactate is certainly produced during short tetani inoxygen (Hill, 1965, pp. 183, 200; von Muralt, 1934),but the amounts are too small to lower the pHdrastically, so that resynthesis would not be muchinhibited by the latter factor. It might therefore be

expected that some of this lactate would be con-verted into glycogen during recovery in 02. Thisindeed may explain why the apparent P/O ratio incontracting frog muscle is so low (Wilkie, 1968).When calculated on the basis of the fact that theheat liberated during recovery in oxygen is aboutequal to the initial contraction heat, and from theknown enthalpies of phosphocreatine hydrolysisand glycogen combustion, it is found that 32mol ofATP is resynthesized/mol of hexose oxidized, andthe P/0 ratio is therefore 2.67 instead of thegenerally accepted value of 3.08.

The authors acknowledge most gratefully the help andadvice given by Dr D. M. Needham during this work andher sustained interest in it. They also thank Dr R. K.Scopes for helpful discussions and for carrying out manyof the enzymic assays. Other analytical work was expertlycarried out by Mr C. C. Ketteridge and Mr J. Holmes, towhom our thanks are also due.

REFERENCES

Bar, U. & Blanchaer, M. C. (1965). Am. J. Phy8iol. 209,905.

Bendall, J. R. & Lawrie, R. A. (1962). J. comp. Path. Ther.72, 118.

Cole, S. W. (1944). Practical Phy8iological Chemi8try,9th ed., p. 167. Cambridge: W. Heffer and Sons.

Eggleton, M. G. & Evans, C. L. (1930). J. Phy8iol., Lond.,70, 269.

Foster, D. L. & Moyle, D. M. (1921). Biochem. J. 15, 672.Good, C. A., Kramer, L. & Somogyi, M. (1933). J. biol.Chem. 100, 485.

Gourley, D. R. H. & Suh, T. K. (1969). Comp. Biochem.Phy8iol. 29, 137.

Hill, A. V. (1965). Trail8 and Trial8 in Phy8iology.London: Edward Arnold Ltd.

Hohorst, J. J. (1963). In Method8 of Enzymatic Analy8i8,pp. 134-266. Ed. by Bergmeyer, H. U. New York:Academic Press Inc.

Krebs, H. A. & Woodford, M. (1965). Biochem. J. 94,436.Mahler, H. R. & Cordes, E. H. (1966). Biological Chemi8try,

p. 441. New York: Harper and Row.Meyerhof, 0. (1920a). Pfluiger Arch. ge8. Phy8iol. 182,313.Meyerhof, 0. (1920b). Pfluigers Arch. ge8. Phy8iol. 182,284.Meyerhof, 0. (1922). Pfltger8 Arch. ge8. Phy8iol. 195, 22.Meyerhof, 0. & Lohmann, K. (1926). Biochem. Z. 171,

421.Meyerhof, O., Lohmann, K. & Meier, A. (1925). Biochem.

Z. 157, 459.Opie, L. H. & Newsholme, E. A. (1967). Biochem. J. 103,

391.Umbreit, W. W., Burris, R. H. & Stauffer, J. F. (1947). In

Manometric Techniques and Related Method8 for theStudy of Ti88ue Metaboli8m, p. 16. Minneapolis:Burgess Publishing Co.

von Muralt, A. (1934). Pfluigers Arch. ge8. Phy8iol. 234,653.

Wilkie, D. R. (1968). J. Phy8iol., Lond., 195, 179.