1148 BIOCHEMISTRY: ROGNSTAD AND KATZ PROC. N. A. S.

"Ibid., 52,1131 (1964).16Geder, L., and A. B. Sabin, unpublished data.17 Herr, R. R., J. Am. Chem. Soc., 81, 2595 (1959).18 Eble, T. E., M. E. Bergy, C. M. Large, R. R. Herr, and W. G. Jackson, Antibiotics Annual,

(New York: Medical Encyclopedia, Inc., 1958-1959), pp. 555-559.19 Haff, R. F., Virology, 22, 430 (1964).20Smith, C. G., W. L. Lummis, and J. E. Grady, Cancer Res., 20, 1394 (1960).

THE BALANCE OF PYRIDINE NUCLEOTIDES AND ATPIN ADIPOSE TISSUE*

BY ROBERT ROGNSTAD AND JOSEPH KATZt

CEDARS-SINAI MEDICAL RESEARCH INSTITUTE, CEDARS-SINAI MEDICAL CENTER,LOS ANGELES, CALIFORNIA

Communicated by Harland G. Wood, February 18, 1966

Determinations of carbon and oxidation-reduction balances are classical tech-niques employed in the investigation of the metabolism of microorganisms. Be-cause of endogenous metabolism, balance studies have rarely been attempted inanimal tissues. However, adipose tissue in vitro is well suited for such studies sincethe catabolism of added glucose greatly exceeds that of endogenous material. Wehave shown' that four products account for over 95 per cent of the glucose utilizedand we have established a complete carbon balance for this tissue. A balance alsohas been established of the formed and utilized cytoplasmic DPNH and TPNH.1-3In the present communication, we present estimations of the formation and utiliza-tion of ATP. The patterns of utilization of energy in epididymal fat pads of ratsunder conditions of high lipogenesis have been found to differ markedly from thoseof high lipolysis.Methods.-Segments, prepared from epididymal fat pads from rats maintained on a fat-free

carbohydrate diet (Nutritional Biochemical Co., Cleveland, Ohio), were incubated in replicatewith glucose-i-C14, 6-C14, and U-C14 (and, in one experiment, with acetate-l-C14 and 2-C14) inthe presence of insulin, epinephrine, or growth hormone. The conditions are described in thelegends to Table 1. The methods of incubation, fractionation of labeled compounds, and isotopicassay have been described previously. At least 95% of the utilized glucose-C14 was accountedfor in CO2, lipid fatty acids, lipid glycerol and lactate, and, under certain conditions (see below),free glycerol. The remainder was recovered in glycogen and other water-insoluble material.

Calculation.-The recoveries of isotope in the products are expressed as specific yields,4 that is,as % of total utilized C14. The C'4 utilized is considered equal to the sum of the C14 recovery in allthe products. The contribution of the pentose cycle was calculated from the specific yields of CO2,fatty acids, and glycerol (from glucose-l-Cl4 and 6-C'4) by two methods presented in detail else-where.", 4-6 (Methods for estimating the contribution of the pentose cycle as well as other calcu-lations are considered in recent reviews.7' 8)

Carbon balance: The conversion of glucose into fatty acids, glycerol, lactate, and CO2 wascalculated from specific yields from glucose-U-C'4. Since incorporation into glycogen was negli-gible, the fraction of utilized glucose converted to triose-P equals [1-(PC/2)].' The fraction oxi-dized to pyruvate was obtained by subtracting the fraction in glycerol from this value. Theamount of pyruvate decarboxylated to CO2 and acetyl-CoA equals this value less the fractionin lactate. The yield of CO2 via phosphogluconate decarboxylation was calculated from thecontribution of the pentose cycle. The yield of CO2 via the Krebs cycle equals the total CO2less the amount derived by the other two pathways described above. Alternatively, the yield of

VOL. 55,1966 BIOCHEMISTRY: ROGNSTAD AND KATZ 1149

CO2 via the Krebs cycle was calculated from the yield of C1402 from glucose-6-C'4.5 The twomethods were in reasonable agreement.

Balance of cytoplasmic pyridine nucleotide: The formation of DPNH by the oxidation of glycer-aldehyde 3-phosphate (GAP), and of TPNH via the pentose cycle was calculated from thecarbon balance. One mole of hydrogen is used in the synthesis of lactate or glycerol and, assum-ing that the fatty acid is palmitate, 1.75 Mtmoles of TPNH per ;mole of acetyl-CoA is required forfatty acid synthesis (2 X 7/8).Results.-The utilization of glucose and specific yields of C'4 are presented in

Table 1. The contribution of the pentose cycle to the metabolism of glucose in thepresence of insulin was 23 per cent when glucose was the sole substrate. It in-creased to 31 per cent when both acetate and glucose served as substrate (expt. B).In the presence of growth hormone and epinephrine, the contribution was 13 percent and 4 per cent, respectively. The use of these two hormones permits observa-tions of major differences in pathways which are observed consistently and re-

TABLE 1GLUCOSE UTILIZATION, SPECIFIC YIELDS, AND ESTIMATED PENTOSE CYCLE IN ADIPOSE TissuE

Experiment: A B C DGrowth

Hormone: Insulin Insulin hormone EpinephrineGlucose (and acetate) utilization 54 63 (7.5) 23 12

;imoles/gm/3 hrSpecific yieldsC02

Glucose-i-C'4 48 60 34 25Glucose-6-C4 2.1 5.2 16 22Glucose-U-C'4 36 40 35 30Acetate-i-C'4 11Acetate-2-C'4 6

Fatty acidGlucose-i-C'4 40 32 17 3Glucose-6-C14 80 82 32 5.4Glucose-U-C'4 49 46 17 4.3Acetate-1-C'4 89Acetate-2-C14 94

GlycerolGlucose-i-C'4 7.5 5.6 23 45Glucose-6-C'4 9.2 8.0 19 34Glucose-U-C'4 9.6 8.7 22 34

LactateGlucose-i-C'4 2.7 1.9 27 26Glucose-6-C14 6.3 4.0 35 38Glucose-U-C14 4.2 4.2 26 31

ResidueGlucose-1-C'4 2.0 0.8Glucose-6-C'4 2.5 1.3Glucose-U-C'4 1.0 1.5

Pentose cycle (%)Method I* 23 31 13 3Method II* 24 31 13 5Epididymal fat pad tissue (300 mg) incubated in 2.5 ml Krebs Henseleit buffer for 3 hr with 25 jsmoles

of glucose (in expt. B with 25 ismoles glucose and 25 pmoles acetate); insulin, 1 unit; growth hormone,4 mg; and epinephrine, 0.1 mg, were added per flask. Specific yields are expressed as per cent of utilizedglucose.

* The pentose cycle (PC) was calculated by two methods, one based on yields in C02 and fatty acidsand the other on yields in glycerol and fatty acids. Both methods were derived from models which takeinto account nonequilibration of triose phosphates.' This is necessary in the case of lipolytic hormones;with insulin-the values agree with those calculated from simpler models.4 6 The square brackets in theequations below indicate specific yields; subscripts indicate the position of labeling of glucose.

Method I PC= St where S' = [CO2]1- y [CO2]6 and = [fatty acidl]M2h"dIIPC^ (1- a) - [glycerol~u(6- e) where di = [fatty acid]65 -( - y) - [glycerol]u(b - -y) glyceol]_Method II: PC =

(+2 y

where I = [glycerol]6

1150 BIOCHEMISTRY: ROGNSTAD AND KATZ PROC. N. A. S.

TABLE 2CARBON BALANCE FROM GLUCOSE METABOLIZED BY ADIPOSE TISSUE

Experiment: A B C D

GrowthHormone: Insulin Insulin hormone Epinephrine

Products:Fatty acid 293 276 103 24Glycerol 58 52 129 204Lactate 25 25 158 186CO2 (pentose cycle) 69 93 39 12CO2 (pyruvate decarb.) 149 143 92 66CO2 (Krebs cycle) 8 21 81 108Fatty acid (from acetate) 70CO2 (from acetate) 6Fatty acid re-esterified* 640 560 1952 3248

Intermediate steps:Hexokinase 600 600 600 600P-fructokinase 462 414 522 576GAP dehydrogenase and pyruvate

kinase 473 455 432 384Acetate activation 0 76 0 0Values are expressed as gatoms carbon per 100 ,umoles of glucose utilized.Example of Calculations for Experiment A.-Hexokinase: Equals glucose utilized (100 ;smoles) and therefore

600 satoms of carbon. Pentose cycle C02: Pentose cycle = 23%; half of the carbon is converted to glycer-aldehydephosphate and half to C02. Hence the C02 from the pentose cycle = 600 (0.23) 1/2 = 69. Phospho-fructokinase: Hexokinase minus pentose cycle = 600 - 2 (69) = 462. Glyceraldehyde phosphate dehydro-genase and pyruvate kinase: Phosphofructokinase plus one-half pentose cycle minus glycerol outflow = 462 +69 - 58 = 473. Pyruvate decarboxylation: Pyruvate kinase minus lactate outflow = 473 - 25 = 448.

448 2One third of this goes to C02 -3 = 149. Two thirds goes to acetyl-CoA = 448 X = 299. Krebs cycleC02: This can be calculated by subtracting the outflow into fatty acids from the outflow into acetyl-CoA =299 - 293 = 6. However, in experiments A and B this was calculated from glucose-6-C'4 (see first footnote).The slight differences in the values calculated by the two methods causes the sum of the outflows to exceed 600slightly in experiments A and B. Fatty acid re-esterification: 58 patoms . 3 = 19.3 pmoles glycerol-P formedfrom glucose; 293 MAatoms - 16 = 18 jsmoles fatty acids (assumed palmitate) formed from glucose; 19.3 jsmolesof glycerol-P will esterify 3 X 19.3 = 58 /Amoles of palmitate; 58 pmoles minus 18 moles (newly synthesizedfatty acids) = 40 pmoles, which is the amount of endogenous fatty acids (produced by lipolysis) that are re-esterified; 40 X 16 = 640 1satoms carbon.

producibly. We therefore have chosen these two extremes as examples for thecalculations.The carbon balance is presented in Table 2. The pattern in the presence of in-

sulin is characterized by high rates of fatty acid synthesis, a significant contribu-tion of the pentose cycle, and a very low contribution of the Krebs cycle. Thepattern in the presence of lipolytic hormones is characterized by a depressed syn-thesis of fatty acid and metabolism by the pentose cycle, and by an increased syn-thesis of glycerol and oxidation by the Krebs cycle.The synthesis of glycerol, under all conditions, was much greater than that re-

quired to esterify the newly synthesized fatty acids. With insulin present, 1 Mmoleof glycerol per 8 Mmoles of acetyl units from glucose was incorporated into lipids[(293/2) *. (58/3) ; 8; Table 2, expt. A]. If the lipid is tripalmitate, 1 Mmole of glyc-erol is sufficient to esterify 3 X 8 = 24 Amoles of acetyl units. Thus, about 2/3 of theesterified fatty acids appear to have been derived from hydrolysis of endogenouslipid. The amount of this re-esterification is calculated in Table 2. In the pres-ence of growth hormone and epinephrine, it exceeds by far the amount of newly syn-thesized acids.The formation and utilization of cytoplasmic-reduced pyridine nucleotides are

presented in Table 3. In the presence of insulin, the calculated yields of TPNH andDPNH almost equal the total equivalents of hydrogen required for reductive syn-thesis of lactate, glycerol, and fatty acids. In the presence of growth hormone,

VOL. 55, 1966 BIOCHEMISTRY: ROGNSTAD AND KATZ 1151

TABLE 3BALANCE OF CYTOPLASMIC-REDUCED PYRIDINE NUCLEOTIDE

Experiment: A B C DGrowth

Hormone: Insulin Insulin hormone EpineplirineFormation (cytoplasm):

Pentose cycle (TPNH) 138 186 78 24GAP dehydrogenase (DPNH) 158 152 144 128

Sum 296 338 222 152Utilization (cytoplasm):

Fatty acid synthesis (TPNH) 256 312 95 21Glycerol (DPNH) 19 18 43 68Lactate (DPNH) 8 8 52 62

Sum 283 338 190 151TPNH "deficit" (F. A. synthesis minus

pentose cycle TPNH) 118 126 17 0Pyruvate dehydrogenase (mitochon-

drial) 149 143 91 6Units are jmoles pyridine nucleotide per 100 ,pmoles glucose utilized.Example of calculations for Expt. A.-Pentose cycle (TPNH) = 3 X 2 X 23 = 138; GAP dehydrogenase

(DPNH) = = 158; fatty acid synthesis (TPNH) =-2 X7X 2 = 256; glycerol synthesis (DPNH) =

58 att ytei DN)=25 473-25= 19; lactate synthesis (DPNH) - = 8; pyruvate dehydrogenase (DPNH) = = 149.3~~~~~~~~~~

there is some excess over the calculated requirement of hydrogen. A possiblereason for this excess is the dilution of the labeled acetyl-CoA by acetyl-CoAformed by the 13-oxidation of fatty acids (produced in large amounts by lipolysis).This would result in an underestimation of the synthesis of fatty acids.TPNH formed in the pentose cycle (Table 3) is not sufficient to provide all the

reducing equivalents for fatty acid synthesis. This has also been observed by Flattand Ball.'0 In agreement with other experiments,1 the maximal contribution ofTPNH from the pentose cycle to reduction of fatty acid ranges from 50 to 60 percent in the presence of insulin to over 80 per cent with lipolytic hormones. Thedeficit of TPNH suggests a transfer of hydrogen from DPNH to TPN. The par-ticipation of DPNH in the synthesis of fatty acid has been established by the useof glucose4-T.2' I

The ATP balance, as calculated from the data of Tables 2 and 3, is presented inTable 4. ATP is produced in the cytoplasm through reactions catalyzed by glycer-aldehyde-3-phosphate dehydrogenase and pyruvate kinase. The yield of ATP inthe mitochondria is readily computed, assuming a yield of 3 ATP in the oxidation ofpyruvate (see below), and a yield of 12 in the oxidation of acetyl-CoA via theKrebs cycle. The requirements of ATP for activation and transport are based onthe following considerations.

(a) Activation of glucose and fructose-6-P: One mole of ATP is required in thehexokinase and phosphofructokinase reaction. (The fraction phosphorylated tofructose-i, 6-diP equals (1-PC) multiplied by the utilized glucose.)

(b) Carboxylation of acetyl-CoA: One mole of ATP is required per mole ofmalonyl CoA formed, and assuming the fatty acids are palmitate, 7/8 mole of ATPper mole of acetyl unit is required for this synthesis.

(c) Transfer of acetyl-CoA: Mitochondria appear to be impermeable to acetyl-CoA. Srere and Bhadurill have proposed a mechanism for transfer to acetyl-CoA (formed in the mitochondria from pyruvate) into the cytoplasm which re-quired 1 mole of ATP per mole of acetyl-CoA." Acetyl-CoA is condensed in the

1152 BIOCHEMISTRY: ROGNSTAD AND KATZ PROC. N. A. S.

TABLE 4ATP BALANCE

Experiment: A B C DGrowth

Hormone: Insulin Insulin hormone EpinephrineUtilization of ATP:

Hexokinase 100 100 100 100Phosphofructokinase 77 69 88 96Acetyl-CoA carboxylase 128 152 47 11Acetyl-CoA transfer* 146 173 54 12Transhydrogenase 118 126 17 0Fatty acid activationt 80 70 244 408Acetate activation 0 76 0 0

Sum 649 766 550 627Formation of ATP:GAP dehydrogenase 158 152 144 128Pyruvate kinase 158 152 144 128Pyruvate dehydrogenase P/0 = 3 448 430 276 198Krebs cycle 48 162 468 648

Sum 812 896 1078 1185Units are lsmoles ATP per 100 ;smoles glucose utilized.Example of Calculations for Expt. A.-Phosphofructokinase = = 77; acetyl-CoA carboxylase = X

2937/8 = 128; acetyl-CoA transfer = = 146; transhydrogenase = 118 (Table 3, TPNH deficit); fatty

acid activation = 6 X 2 = 80; GAP dehydrogenase = pyruvate kinase = 158 = DPNH yield from GAP

dehydrogenase (Table 3); pyruvate dehydrogenase =4

X 3 = 448; Krebs cycle = 2 X 12 = 48.3 ~~~2 -* In expt. B, acetate activation is assumed to occur intramitochondrially; this is unproved.t Prior to re-esterification of free fatty acids produced by lipolysis.

mitochondria with oxalacetate to form citrate, which passes into the cytoplasm.There, citrate is split to acetyl-CoA and oxalacetate by the citrate cleavage enzyme,a reaction requiring ATP.

(d) DPNH-TPN transhydrogenase: Since the ratio DPNH/DPN+ in cells ismuch smaller than the ratio TPNH/TPN+, the transfer of hydrogen from DPNHto TPN will require energy. A scheme for such hydrogen transfer was proposed byLowenstein.'2 The reactions are catalyzed by pyruvate carboxylase, malate de-hydrogenase, and malate enzyme as follows:

pyruvate carboxylaseATP + pyruvate + CO2 - ) oxalacetate + ADP + Pi

malate dehydrogenaseOxalacetate + DPN+ - Inalate + DPNH + H+

malate enzymeMalate + TPN+ -4 pyruvate + CO2 + TPNH + H+

Net: DPNH + TPN + ATP DPN+ + TPNH +ADP + Pi

The ATP required for this transfer was calculated from the TPNH deficit (bottomrow of Table 3).

(e) Activation of fatty acid: The activation of a fatty acid from endogenouslipid as well as the activation of exogenous acetate involves the formation of 1 moleof pyrophosphate per mole of acid and thus requires 2 moles of ATP for resynthesisof ATP.Table 4 indicates that in the presence of insulin the amount of ATP utilized in the

synthetic reactions accounts for the major share of the calculated ATP formed.On the other hand, in the presence of lipolytic hormones, the calculated formation

VOL. 55, 1966 BIOCHEMISTRY: ROGNSTAD AND KATZ 1153

of ATP is nearly twice that required for synthesis. Actually, under these conditionsthe capacity for ATP production is probably somewhat higher, since oxidation ofendogenous fatty acids is likely to occur. Flatt and Ball have reported that oxida-tion of endogenous fatty acid is small in the presence of insulin and absence ofepinephrine.10Discussion.-Compartmentalization of reduced pyridine nucleotides: Under widely

different experimental conditions, as shown previously" I and in Table 3, theequivalents of hydrogen supplied by the pentose cycle and by oxidation of glyceralde-hyde phosphate are sufficient for the synthesis of glycerol, fatty acids, and lactate.All these reactions occur in the cytoplasm. We have shown, with the use of glu-cose-3-T, that all the TPNH generated in the pentose cycle is recovered in fattyacids.' 14 We have also shown that hydrogen from position 4 of glucose, which istransferred to DPNH, appears in glycerol and lactate, and also in the presence ofinsulin provides from one quarter to one half of the equivalents of hydrogen re-quired for synthesis of fatty acid.' Apparently extramitochondrial oxidation ofreduced pyridine nucleotide does not occur to an appreciable extent (for example, bya microsomal cytochrome system). It also appears that transfer of reducing equiva-lents between the cytoplasm and mitochondria is either absent or slight."3

Carefully prepared mitochondria are impermeable to reduced pyridine nucleo-tides, and experimental evidence has been furnished for separate intra- and extra-mitochondrial pools of DPNH in the intact cell.'8 However, several mechanismshave been proposed for the transfer of hydrogen between the compartments. Thedihydroxyacetone phosphate-glycerophosphate shuttle could transfer reducingequivalents from the cytoplasm to the mitochondria, and it has been suggested tofunction in this manner in flight muscle.'9 Transfer of hydrogen out of the mito-chondria might also occur (possibly via a system involving acetoacetate and j3-hydroxybutyrate), whereby DPNH produced intramitochondrially (via DPN-linked enzymes, or flavin-linked enzymes, and "reverse oxidative phosphoryl-ation"20) could be used for synthesis in the cytoplasm. Transport of citrate out ofthe mitochondria and subsequent oxidation via TPN isocitrate dehydrogenase isalso a possible mechanism by which cytoplasmic TPNH might be formed at the ex-pense of mitochondrial DPNH. Our data on adipose tissue metabolism indicatethat such systems for hydrogen transfer are either absent or operate to a verylimited extent.

The pyruvate cycle: The pyruvate cycle may be pictured as follows (subscriptsm = mitochondria, c = cytoplasm):

Pyruvatem + ATP + CO2 oxalacetatem + ADP + PiAcetyl-CoAm + oxalacetatem citrate,, + CoA

Citratem - citrate,Citrate, + ATP + CoA oxalacetate, + acetyl-CoA, + ADP + Pi

Oxalacetate, + DPNHC + H+ - malate, + DPNC+Malate, + TPNC -- pyruvate, + TPNHC + H+ + CO2

Net: Acetyl-CoA_ + 2ATP + DPNHC + TPNC -- acetyl-CoAc + 2ADP +2Pi+ DPNc + TPNHc

1154 BIOCHEMISTRY: ROGNSTAD AND KATZ PROC. N. A. S.

It is seen that the transhydrogenase and the acetyl-transfer systems both involvepyruvate and oxalacetate and the reactions overlap and thus form a pyruvate cycle.There is good experimental support for the operation of a transhydrogenase sys-

tem involving this cycle. The activity of cytoplasmic malate dehydrogenase isvery high in adipose tissue.2" Pande et al.22 demonstrated a transfer of DPNH toTPN in adipose tissue via cytoplasmic malate dehydrogenase and the malate en-zyme. Young et al.23 showed that the effect of diet on the activity of malate enzymeparalleled the stimulatory effect on lipogenesis. Wise and Ball'6 demonstratedpyruvate carboxylase in adipose tissue. The activity of the citrate cleavage enzymehas also been shown to parallel that of fatty acid synthesis in liver and adipose tis-sue.24, 25Under some conditions most of the TPNH for fatty acid synthesis is supplied by

the pentose cycle (see Table 3), and in others more than half of the TPNH is sup-plied by the trapshydrogenase (thyroxin-treated rats, Katz and Giveon, unpub-lished). In the former circumstances the oxalacetate resulting from the transfer ofacetyl CoA must be removed, possibly by decarboxylation. While the details ofthe pyruvate cycle require elucidation, the evidence for this scheme seems strong.The interaction of the pyruvate cycle and the Krebs cycle may account for therandomization of C14 in lactate from glucose-6-Cl4 which is observed in adiposetissue.' The pyruvate cycle is quantitatively a major metabolic system in adiposetissue and mammary gland (Katz, Giveon, and Rognstad, unpublished), whilemetabolic flow along the Krebs cycle is minor under conditions of rapid synthesis offatty acids.ATP balance: Experiments A and B, Table 4, show that in the presence of insu-

lin most of the ATP produced (assuming the conventionally accepted P/O ratiosand tightly coupled oxidative phosphorylation) was utilized for activation re-actions and transport. The P/O ratio for the pyruvate dehydrogenase step isgenerally believed to be 3, although the experimental evidence is perhaps not aswell established as for other steps of the Krebs cycle.26 Probably other functionsrequiring ATP, such as transport, will account for the calculated excess. Studiesof the energy metabolism of bacteria have indicated that the calculated productionof ATP generally exceeds by threefold the estimation of ATP required for syntheticpurposes.27 Here the variety of synthetic reactions is much greater; also, theamount of energy used for bacterial motility and organization must be considered.

In the presence of lipolytic hormones (expts. C and D), if efficient oxidative phos-phorylation were assumed, much more ATP would be formed than required. Thisfact suggests that' uncoupling must be extensive. Such uncoupling could be causedby the high concentration of free fatty acids produced by the rapid lipolysis. Fattyacid activation required for re-esterification is the largest single drain of ATP inthese experiments. It is possible that this uncoupling is related to the proposedfunction of adipose tissue as a heat-generating organ.21The balance studies presented here are based on a number of assumptions, the

validity of which is yet to be established. The estimates can be made more reliableby including simultaneous measurements of oxygen uptake, formation of free fattyacid, and release of glycerol. Such studies are in progress.Summary.-In vitro studies with rat adipose tissue indicate an almost equivalent

balance between formation and utilization of reduced pyridine nucleotides in the

Vol.. 55, 1966 BIOCHEMISTRY: ROGNSTAD AND KATZ 1155

cytoplasm. Estimations of the metabolism of ATP indicate that requirements forknown synthetic reactions account for the majority of the ATP produced in thepresence of insulin. In the presence of lipolytic hormones, the amount of ATPwhich could be generated is nearly twice that required for synthetic reactions, sug-gesting extensive uncoupling.

* Supported by grant no. AM03682-06 from the National Institutes of Health and grant-in-aid no. 64 G 81 from the American Heart Association.

t Work done during Established Investigatorship of the American Heart Association.1 Katz, J., B. R. Landau, and G. E. Bartsch, J. Biol. Chem., 241, 727 (1966).2 Rognstad, R., and J. Katz, Federation Proc., 24, 591 (1965).3 Katz, J., and R. Rognstad, J. Biol. Chem., July 1966.4 Katz, J., and H. G. Wood, J. Biol. Chem., 238, 517 (1963).6 Wood, H. G., and J. Katz, J. Biol. Chem., 233, 1279 (1958).6 Katz, J., and H. G. Wood, J. Biol. Chem., 235, 2165 (1960).7 Wood, H. G., J. Katz, and B. R. Landau, Biochem. Z., 338, 809 (1963).8 Landau, B. R., and J. Katz, in Handbook of Physiology: Section 5, ed. A. E. Renold and

G. F. Cahill, Jr. (Washington, D.C.: American Physiological Society, 1965), p. 253.9 The specific yield of C'402 from glucose-6-C'4 was divided by 1.20, the approximate relative

specific activity of glyceraldehyde phosphate from glucose-6-C'4,', 8 and multiplied by 2/3 (sinceonly 2/3 of the carbon of triose phosphate is converted to acetyl-CoA) and by a factor 2/(1 + 1/2).The latter factor was to correct for unequal oxidation of the two carbons of acetyl-CoA, as evi-denced by the ratio of 2 for C1402 from acetate-i-C14 and 2-C14 (Table 1).

'0Flatt, J. P., and E. G. Ball, J. Biol. Chem., 239, 675 (1964).11 Srere, P., and A. Bhaduri, Biochim. Biophys. Acta, 59, 487 (1962).12 Lowenstein, J. M., J. Biol. Chem., 236, 1213 (1961).13 Flatt and Ball suggested that the pyruvate dehydrogenase system may be extramitochondrial

and that the reducing equivalents from this reaction may be available for synthesis of fatty acids.'0The suggestion was apparently based on the finding that in adipose tissue from fasted rats thesynthesis of fatty acids from acetate (in the absence of glucose) was stimulated by pyruvate."6We performed an experiment under similar conditions and also found a stimulation by pyruvate.However, succinate stimulated synthesis to a greater extent. Since succinic dehydrogenase isdefinitely mitochondrial, it is likely that the stimulating effect of pyruvate and succinate is in-direct. Under the conditions of the experiment fatty acid synthesis is very low, compared to syn-thesis from glucose in adipose tissue of fed animals, and TPN-dependent isocitrate dehydrogenasemight furnish the required reducing equivalents. If pyruvate dehydrogenase were extramito-chondrial, a large excess of reduced pyridine nucleotide over synthetic needs would be available(Table 3). Ball and co-workers'0, 16, 17 have suggested that reoxidation by oxygen of this sup-posed surplus may be a limiting step in lipogenesis. Our results support the view that no oxida-tion (with 02) of cytoplasmic pyridine nucleotide occurs.

14 Katz, J., R. Rognstad, and R. Kemp, J. Biol. Chem., 240, PC1484 (1965).15 Winegrad, A. I., and A. E. Renold, J. Biol. Chem., 233, 267 (1958)."f Wise, E. M., Jr., and E. G. Ball, these PROCEEDINGS, 52, 1255 (1964).17 Flatt, J. P., and E. G. Ball, in Handbook of Physiology: Section 5, ed. A. E. Renold and G. F.

Cahill, Jr. (Washington, D. C.: American Physiological Society, 1965), p. 273.18 Chance, B., and B. Thorell, J. Biol. Chem., 234, 3044 (1959).19 Sacktor, B., and A. Dick, J. Biol. Chem., 237, 3259 (1962)." Azzone, G. F., L. Ernster, and E. G. Weinbach, J. Biol. Chem., 238, 182 (1963).21 Durr, I. F., and B. Dajani, Comp. Biochem. Physiol., 13, 225 (1964).22 Pande, S. V., R. P. Khan, and T. A. Venkitasubramanian, Biochim. Biophys. Acta, 84, 239

(1964).23 Young, J. W., E. Shrago, and H. A. Lardy, Biochemistry, 3, 1961 (1964).24 Kornacker, M. S., and J. M. Lowenstein, Biochem. J., 94, 209 (1965).25Kornacker, M. S., and E. G. Ball, these PROCEEDINGS, 54, 899 (1965).2 Kornacker and Ball have proposed2' that the P/O ratio for the pyruvate dehydrogenase reac-

tion may be 2, since a flavin is involved in the system. If this P/O ratio is assumed, a nearly

1156 BIOCHEMISTRY: ROSEN AND ROSEN PROC. N. A. S.

perfect balance would result between ATP formation and utilization in experiments A and B ofthis paper. This, however, is probably fortuitous.

27 Gunsalus, I. C., and C. W. Shuster, in The Bacteria, ed. I. C. Gunsalus and R. Y. Stanier(New York: Academic Press, 1961), vol. 2, p. 46.

28 Cahill, G. F., Jr., in Adipose Tissue as an Organ, ed. L. W. Kinsell (Springfield, Illinois:Charles C Thomas, 1962), p. 126.

CHEMICAL MODIFICATION OF THE ALLOSTERIC ANDCATALYTIC SITES OF FRUCTOSE 1,6-DIPHOSPHATASE*

By ORA M. ROSENt AND SAMUEL M. ROSENtDEPARTMENT OF MEDICINE, ALBERT EINSTEIN COLLEGE OF MEDICINE, BRONX, NEW YORK

Communicated by B. L. Horecker, March 7, 1966

Crystalline fructose 1,6-diphosphatase (FDPase)l isolated from Candida utilis2has been shown to have properties generally similar to those of FDPases isolatedfrom other sources.3 The enzyme is highly specific for the substrate FDP, which isconverted to F6P; it has a molecular weight of approximately 100,000 and it ex-hibits an alkaline pH optimum (pH 9.0-9.5). In the presence of low concentrationsof EDTA, however, a new pH optimum at pH 7.5-8.0 is observed.

Since FDPase catalyzes a critical and essentially irreversible step in gluconeo-genesis, its regulation in vivo is of particular interest. Most FDPases are inhibitedby AMP,4 5 and it has been suggested that the intracellular concentration of AMP isan important factor in the control of FDPase activity in vivo. We have previouslyshown that 5'-AMP is a specific inhibitor of the enzymic activity induced by EDTAat pH 7.5.2 The inhibition is reversible and noncompetitive with substrate. Noinhibition is observed with adenosine, Pi, or other nucleotides including 2'-AMP,3'-AMP, and 3', 5'-cyclic AMP. The suggestion that AMP may play a role in thephysiological control of glycolysis is supported by the observation that phospho-fructokinase, which catalyzes the opposing reaction, i.e., the formation of FDPfrom F6P, is activated by 5'-AMP.6 Thus, a rise in the concentration of intracellu-lar AMP would enhance phosphofructokinase activity, prevent the breakdown ofFDP to F6P, and thereby promote glycolysis. Conversely, a decrease in the con-centration of AMP would release FDPase from inhibition and allow metabolismto proceed in the direction of gluconeogenesis.7We now have found that the inhibitory effect of 5'-AMP can be completely abol-

ished by treating FDPase with FDNB in the presence of substrate. Under theseconditions there is no loss of catalytic activity. Dinitrophenylation in the absenceof FDP, however, leads to inactivation. A similar but reversible desensitization toAMP inhibition is produced by urea or PMB. The use of C14-FDNB has permittedthe qualitative and quantitative evaluation of the amino acids involved in bothcatalytic activity and desensitization to the allosteric modifier, AMP.8

Materials and Methods.-FDPase was purified and recrystallized twice from commerciallygrown C. utilis by a modification of the previously published procedure.2 The purified enzymecatalyzed the conversion of 80 ,umoles of FDP to F6P per min per mg protein. Enzyme assayswere performed at 250 with a Beckman DU spectrophotometer attached to a Gilford multiple-