INTERRELATION OF STROMAL NAD(P)ASE AND HUMANERYTHROCYTIC 6-PHOSPHOGLUCONIC DEHYDROGENASE*

BY FRANCO AJMAR, tBARBARA SCHARRER, FRED HASHIMOTO, ANDPAUL E. CARSONT

SECTION OF GENETICS AND INTERNAT1ONAL HEALTH, DEPARTMENT OF MEDICINE,UNIVERSITY OF CHICAGO, CHICAGO, ILLINOIS

Communicated by Leon 0. Jacobson, December 21, 1967

Pentose phosphate metabolism of human red blood cells may be altered byinherited deficiency in cytoplasmic enzymes like glucose-6-phosphate dehydro-genase (D-glucose-6-phosphate:NADP oxidoreductase, E.C. 1.1.1.49), result-ing in increased hemolytic susceptibility.' This deficiency may be due either to amutation at a structural gene or to an alteration in the rate of synthesis or ofdestruction of the enzyme. Two electrophoretic variants of glucose-6-phosphatedehydrogenase not accompanied by enzyme deficiency have been demonstratedto differ by a single amino acid substitution, and therefore a mutation at a struc-tural gene has been postulated.2 On the other hand, interaction of stromalNAD(P)ase (NAD(P) glycohydrolase, E.C. 3.2.2.6) with glucose-6-phosphatedehydrogenase has been described as a possible mechanism for regula-tion of the stability of the enzyme.3' 4 Furthermore, two other cytoplas-mic enzymes, glutathione reductase (reduced NAD(P):oxidized glutathioneoxidoreductase, E.C. 1.6.4.2) and 6-phosphogluconic dehydrogenase (6-phospho-D-gluconate: NAD(P) oxidoreductase, E.C. 1.1.1.44) can be modified in theiractivity by incubation of hemolysates in the presence of red cell membranes(stromata), suggesting a more general regulatory function of the membranes onthe cytoplasmic enzymes.5 6 These modifications consist of activation of gluta-thione reductase and, when NADP is added to the incubation mixture, inactiva-tion of 6-phosphogluconic dehydrogenase accompanied by change in electro-phoretic pattern.' This last effect of the membranes has now been further inves-tigated because of its specificity in the requirement of the coenzyme and becauseit was accompanied by a structural change in the enzyme. The present dataprovide evidence that (1) the stromal factor required to induce this effect on6-phosphogluconic dehydrogenase is NAD(P)ase and (2) a product of theNAD(P)ase reaction with NADP, P-ADPR, interacts with 6-PGD to cause itsmolecular alteration.Methods.-Freshly drawn heparinized human blood was used in all our experiments.

After the washed red cells had been frozen and thawed, stroma-free hemolysate was pre-pared by centrifugation as previously described6 and, in addition, was passed through aMillipore filter (pore size, 0.80 1A) and checked with a particle counter (Coulter Counter B)for absence of stromata. In some experiments, stroma-free hemolysate was dialyzed bygel filtration through Sephadex G25 eluted with 0.162 M NaCI, and is referred to asdialyzed hemolysate. Partial purification of 6-PGD was accomplished by gel filtrationthrough Sephadex G200 eluted with 0.05 M Tris-HCl buffer and 0.1 M KCl, pH 7.4.The fractions containing 6-PGD activity collected with a Gilson fraction collector formeda peak preceding hemoglobin and are referred to as purified 6-PGD.The activity of 6-PGD was measured at 370C by the method of Glock and McLean.8

Unless otherwise stated, the data refer to the most common human variant, Pd A, asdetermined by starch gel electrophoresis.9

538

Dow

nloa

ded

by g

uest

on

June

1, 2

020

BIOCHEMISTRY: AJMAR ET AL.

Hemoglobin-free stromata were stored overnight at 40C after being prepared essentiallyas described by Dodge et al.,lo with the modification that the osmolalities of the variouswashing buffers were monitored during the preparation with anl osmometer (Mechrolabvapor pressure osmometer), rather than estimated by calculation. The recovery andnumber of stromata were determined with the particle counter.

Vertical starch gel electrophoresis and staining of 6-PGD were accomplished accordingto Bowman et al.9 except that Electrostarch (Electrostarch Co., Madison, Wisconsin)(52.5 gm/500 ml of Tris borate buffer) was used in the preparation of the gel. Ascendingpaper chromatography was performed as described by Zatman et al." Inorganic phos-phorus was measured according to the method of Sumner.12 For pentose determinationthe method of Ashwell" as modified by Kellermeyer et al.'4 was used. Adenosine wasestimated spectrophotometrically at 260 mu using a molar extinction coefficient of 1.54 X104.15The NAD(P)ase activity of hemoglobin-free stromata was assayed with the KCN

method of Colowick et al.'" The units are expressed as Moles of NADP split per hourper 109 stromata. The system contained 2.5 X 109 stromata per ml and NADP at aconcentration of 7.7 X 10-4M. (All preparations showing NADPase activity used inthis investigation also had NADase activity approximately twice that of the NADPase.To our knowledge, these activities have not been separated biochemically, but there isindirect evidence suggesting that these may be two enzymes.'7 Good agreement was ob-tained by this method with the results obtained by enzymatic assay with glucose-6-phos-phate dehydrogenase (Sigma Chemical Co.) for NADP, and alcohol dehydrogenase(Sigma) for NAD.)Experimental and Results.-Stromal effect on 6-PGD was obtained by incubat-

ing hemoglobin-free stromata and NADP with dialyzed hemolysate or purified6-PGD at 450C for two hours unless otherwise specified. The complete systemscontained a final count of 2 X 109 stromata/ml and a final concentration of 1 X10-4 M NADP. The stromata were then removed by centrifugation, and ali-quots of the supernatant were taken for assay of the 6-PGD activity and forstarch gel electrophoresis. The effects on the electrophoretic pattern are pre-sented in Figure 1. Instead of a smearing of the electrophoretic pattern, as

FIG. .--Starch gel electrophoresis ofpurified 6-PGD showing the changes inpattern induced by incubation with hemo- fo kglobin-free stromata, and NADP at 370C.0See text for preparation of the incubationsystems. Slot 1, nonincubated fraction;slots 2-6, stromata + NADP + fractionafter incubation at 370C for 0.5, 1, 2, 4,and 7.5 hr, respectively; slot 7, fractionalone after 4 hr incubation; slot 8, frac- 4,tion + NADP after 4 hr incubation; ,-Bslots 9 and 10, fraction + stromata after4 hr and 2 hr incubation, respectively. 1 2 3 4 - 6 7 8 9 10Entirely comparable results are obtained 3on incubation at 451C but the process iscomplete in 2 hr.

previously reported,7 there is a progressive disappearance of the leading band withconcomitant appearance of a slower band. The slower band continues to migrateahead of the first band, which gradually stains more intensely but does not changein mobility. The final pattern is quite similar to that of the relatively rare

homozygous Pd B enzyme. The pH optimum of the Pd A enzyme modified by

VOL. .59, 1968 539

Dow

nloa

ded

by g

uest

on

June

1, 2

020

BIOCHEMISTRY: AJMAR ET AL.

stromal effect does not change, however, and is different from the pH optimum ofthe Pd B enzyme at 370C (Fig. 2).

80-

70-

it40 -PdA-A0--0 Pd B-0--- Pd A modified

20

TO 72 4 76 78 8D 92 8.4 8.6 98 9D 92 94pH

FIa. 2.-pH optima curves at 370C of three genetic formsof 6-JPGD (Pd A, Pd A-B, and Pd B) and of Pd A modified bystromal effect. Tris buffer (I = 0.1) at different pH's wasused in the assay system. The activity is expressed as per-centage of the maximum activity.

Since both NADP and stromata were specifically required to induce thiseffect,7 and stromata contain NAD(P)ase,4 17,15 the possible relationship ofstromal effect to NAD(P)ase was examined and three correlations were studied:

(1) Correlation with intrinsic NAD(P)ase activity of stromata: We have con-firmed the relative deficiency of NADase in stromata of the blood of newborns,19and have shown that they are also relatively deficient in NADPase activity.In addition, stromata of the blood from three adult subjects exhibited lowNAD(P)ase activities (about 25% of the normal). Hemoglobin-free stromata oflow NAD(P)ase activity, when incubated with stroma-free hemolysate in the

TABLE 1. Rates of inactivation of 6-PGD by stromata with different NAD(P)aseactitity.

-6-PGD Activity After Various Lengths of Incubation at 45C-0 15 30 60 120 240

Sample no. minmmrin min ein min1 0.741 0.749 0.704 0.644 0.576 0.421

(100.0) (101.1) (95.0) (86.9) (77.7) (56.8)2 0.747 0.695 0.638 0.541 0.371 0.167

(100.0) (93.0) (85.4) (72.4) (49.7) (22.4)

The incubation mixtures contained stroma-free hemolysate at a final hemoglpbin concentra-tion of 3.75 gm/100 ml, 10-4 M NADP and (1) stromata (1.2 X 109/ml) with "low NAD(P)ase"activity (0.01 units) in sample 1; (2) stromata (0.9 X 109/ml) with "normal NAD(P)ase" activity(0.1 units) in sample 2.The activity is expressed as moles X 10-8 ml-' min-' of reduced NADP referred to a final

hemoglobin concentration of 0.105 gm/100 ml.In parentheses are the values of 6-PGD activity as per cent of the value at zero time.

540 PROC. N. A. S.

Dow

nloa

ded

by g

uest

on

June

1, 2

020

BIOCHEMISTRY: AJMAR ET AL.

presence of NADP, showed a slow rate of inactivation of 6-PGD compared tothat found with stromata of normal NAD(P)ase activity incubated with thesame stroma-free hemolysate (Table 1). The relative rate of change of the elec-trophoretic pattern was in accord with these results (Fig. 3). Cross experimentswith stroma-free hemolysate from blood with low stromal NAD(P)ase activity

11b

* A.~~~A

1 23 4

1 2 3 4 5 6 7 8 9 10

FIG. 4.-{;PhosphogluconicFIG. 3.-6-Phosphogluconic dehydrogenase dehydrogenase patterns on

patterns after starch gel electrophoresis of the starch gel electrophoresis of thesamples of Table 1. Slots from 1 to 5 correspond incubation mixtures describedto sample 1(15, 30, 60, 120, and 240 min). Slots in Table 2. Slot number cor-from 6 to 10 correspond to sample 2 (240, 120, responds to sample number.60, 30, and 15 min). Hb = hemoglobin.

incubated with stromata of normal NAD(P)ase activity resulted in a normal rateof inactivation.

(2) Correlation with nicotinamide inhibition of NAD(P)ase: NAD(P)aseactivity of hemoglobin-free stromata was completely inhibited by addition ofnicotinamide at a final concentration of 1 X 10-1 M. The same concentration ofnicotinamide completely prevented stromal effect on 6-PGD both by assay andby electrophoresis. When successively lower concentrations of nicotinamidewere used (10-2 to 10' M), there was proportionately less inhibition of stro-mal effect on the starch gel.

(3) Correlation uwth NAD(P)asefrom a different 8ource: NAD(P)ase (50 units)from Neuro8pora crassa (Worthington Biochemical Corp.) added to stroma-freehemolysate in the presence of 10-4 M NADP induced the electrophoretic change.In addition, Neurospora 6-PGD, present in this commercial preparation ofNAD(P)ase as a contaminant with different electrophoretic mobility, wasalso modified in the same incubation mixture.Having established a relationship between NAD(P)ase activity and stromal

effect, the two reactions were then examined consecutively. First, hemoglobin-free stromata were incubated with or without 10-4 M NADP for two hours at450C. The stromata were then removed by centrifugation, and the supernatants,either boiled for three minutes at 1000C or not boiled, were added to dialyzedhemolysate with 10-4 M NADP and incubated at 450C for two hours. Onlythe supernatants collected after incubation of stromata with NADP induced thechange in 6-PGD, both by assay and by electrophoresis (Table 2, samples 3 and4; Fig. 4, slots 3 and 4.) Furthermore, when aouspension of hemoglobin-free

VOL. 59, 196 MA41

Dow

nloa

ded

by g

uest

on

June

1, 2

020

BIOCHEMISTRY: AJMAR ET AL.

stromata and 10-4M NADP in 0.007 M phosphate buffer (pH 7.4) was placed ina dialysis bag and the bag was suspended in dialyzed hemolysate, a stromal effectwas induced on the hemolysate 6-PGD after incubation at 450C for two hours.No stroinal effect was induced when a suspension of stromata in phosphate bufferwithout NADP was used.

TABLE 2. Inactivation of 6-PGD by thermostable products of NAD(P)ase reaction.Sample no. 6-PGD activity Sample no. 6-PGD activity

1 0.545 3 0.3782 0.552 4 0.346

Stromata (2 X 109/10 ml) were incubated with 0.007 M phosphate buffer pH 7.4 (samples 1and 2) or with 10-4 M NADP in phosphate buffer pH 7.4 (samples 3 and 4) for 2 hr at 450C, thenremoved by centrifugation. The supernatants of samples 1 and 3 were boiled for 3 min at 1000C;aliquots of dialyzed hemolysate (hemoglobin concentration = 2.02 gm/100 ml), incubated for 2 hrat 450C with each of the four supernatants and with 10-1 M NADP, were tested for 6-PGD activ-ity.

These results suggested that a heat-stable dialyzable product of the reaction ofNAD(P)ase with NADP was responsible for the stromal effect on 6-PGD. Toisolate the components of the NAD(P)ase reaction and study their effect on6-PGD, NAD(P)ase (100 units) from Neurospora crassa was incubated at 450Cin the presence of 10-2 M NADP. After 60 minutes, the mixture was boiled for

2.0 ' M " 1

1.0 0 5404 0 56062

a~~~~~~~~~~~~40~~~

1.5 -r

cN 0Z

pekIII =-APR

8 .0

(Amicon~~mCroainadtepoutofteratowrech-oat)ape

photometric 0mesrmn fte ltdmtra.)6 rvaedtredsic

0.5-~~~~~~~~~~.10 Z

20 25 30 35 40 45 50 55 60 65 70TUB6E NUMBER

FIG. 5.-DEAE Sephadex A25 chromatography of theproducts of the reaction between NAD(Paese and NADP(see text). Peak I = nicotinamide; peak II = NADP;peak III = P-ADPR.

10 minutes, the proteins were removed by ultrafiltration with a Diaflo ultrafliter(Amicon Corporation), and the products of the reaction were chromatographedthrough a DEAE Sephadex A25 column, with a linear NaCl gradient. Spectro-photometric measurement of the eluted material at 260 mju revealed three distinctpeaks (Fig. 5). The material of each peak was detected as a single spot on paperchromatography. Nicotinamide (Rf = 0.802) 'was found to constitute one peakand NADP (Rf = 0.523) another. The component of the third peak migratedbetween nicotinamide and NADP, with an Rf of 0.685. For the third peak, the

542 PRoc. N. A. S.

Dow

nloa

ded

by g

uest

on

June

1, 2

020

BIOCHEMISTRY: AJMAR ET AL.

TABLE 3. Inactivation of 6-PGD by P-ADPR.-6-PGD Activity after Incubation for t Minutes-

at 450CSample t = 0 60 120

1. Nicotinamide2. NADP3. P-ADPR(X1)4. P-ADPR (X l/lo)

0.7170.7370.7270.750

0.6850.6870.4290.595

0.6600.6560.2810.515

6-Phosphogluconic dehydrogenase activities after incubation of dialyzed hemoly-sate (final hemoglobin concentration 1.5 gm/100 ml) and the components of theNAD(P)ase reaction with NADP shown in Fig. 5.: Sample 1 = nicotinamide (peakI); sample 2 = NADP (peak II); sample 3 = P-ADPR (peak III); sample 4 =P-ADPR diluted 1:10.

ratio adenine: ribose: phosphorus was 1:2.25: 3.00. Since NAD(P)ase specificallycleaves NADP at the nicotinamide-ribose link,10 this finding is consistent withthe assumption that the product of the third peak is P-ADPR. The material ofeach peak was incubated for different times at 450Cwith dialyzed hemolysate. Only the components ofpeak 3 induced a stromal effect on the 6-PGD bothby assay and by electrophoresis (Table 3, sample 3;Figure 6, slot 3). P-ADPR was also found to berelatively specific since adenosine diphosphate Jribose, 5'-AMP, 2'-AMP, 3'-AMP, 5'-IMP, and 0adenosine, incubated with dialyzed hemolysate, did 1 2 3 4not induce stromal effect.Discussion.-We had previously recognized (1) FIG. 6.-Electrophoretic

that neither free nor bound NADP was necessary pattern of 6-PGD of thefor catalytic stability of 6-PGD,7 (2) that NADP incubation mixtures described

in Table 3 after 120 min.was nevertheless related to the structure of 6-PGD Slot numbers correspond towhen stromata were present,7 and (3) that an as yet sample numbers.unknown product might be responsible for the stro-mal effect on 6-PGD.6 The present data allow qualitative reconciliation ofthese observations as follows:

(1) NADP stromal NAD(P) ase P-ADPR

(2) P-ADPR + 6-PGD -* 6-PGD mod.

where 6-PGD mod. represents the structurally and functionally modified enzyme.This indirect action of NAD(P)ase on 6-PGD contrasts with the direct mecha-nism of inactivation of glucose-6-phosphate dehydrogenase by stromal NAD-(P)ase,4 which destroys the NADP required for stability of the enzyme.2'NADP protects G-6-PD against inactivation by NAD(P)ase, even while par-ticipating in the structural modification of 6-PGD. As might be expected, there-fore, incubation of hemolysates with P-ADPR does not result in inactivation ofthe G-6-PD in these hemolysates. Ben-Hayyim et al.,'5 however, have recentlyreported competitive inhibition of some NADP-dependent systems by P-ADPR,although they did not test 6-PGD. Whether NADP itself is involved in thestructure of 6-PGD is iiot known. Although the coenzyme is not required in the

VOL. 59, 1968 543

Dow

nloa

ded

by g

uest

on

June

1, 2

020

BIOCHEMISTRY: AJMAR ET AL.

purification steps of Candida utilis,22 its use in starch gel electrophoresis preventsdiffusion of the bands,9 and its omission has allowed the description of a newmutant.23

Further investigation of the reactions of P-ADPR with 6-PGD from varioussources should be useful in elucidating the structure and genetic regulation ofthe enzyme. Two genetic variants of 6-PGD, Pd A-B and Pd B, although similarto 6-PGD mod. in electrophoretic pattern, differ from it in at least one biochem-ical parameter, i.e., pH optimum. This result suggests that the difference inelectrophoretic pattern among the natural variants is the result of a mutation ata structural gene for 6-PGD and not the effect of an altered stromal NAD(P)aseactivity in intact cells. Regulation of 6-PGD activity by the NAD(P)ase-NAD(P) reaction may be effective, however, in the intact anucleated red cell or inother nucleated normal (or tumor) cells, where microsomal NAD(P)ase acts onoxidized coenzymes present in the cytoplasm or released from the swelling of themitochondria.24Summary.-Red cell membranes induce inactivation and change in structure

of hemolysate 6-phosphogluconic dehydrogenase in vitro in the presence of NADP.NAD(P)ase, present in the red cell membranes, splits NADP, and one of theproducts, 2'-phosphoadenosine diphosphate ribose, modifies the structure andthe activity of 6-phosphogluconic dehydrogenase.

Abbreviations: NADP, NADPase: nicotinamide-adenine dinucleotide phosphate andNADP nucleosidase; NAD, NADase; nicotinamide-adenine dinucleotide and NAD nucleo-sidase; NAD(P): NAD, or NADP; P-ADPR; 2'-phosphoadenosine diphosphate ribose;6-PGD; 6-phosphogluconic dehydrogenase; G-6-PD: glucose--phosphate dehydrogenase;5'-,2'-,3'-AMP: 5'-, 2'-, 3'-phosphates of ribosylnucleosides of adenine; 5'-IMP: inosine-5'-phosphate.

* Contribution number 286 from the Army Research Program on Malaria. This work wassupported by U.S. Public Health Service grant HE 06078, and by U.S. Army contracts DA-49-193-MD-2413 and DA-49-007-968. Taken in part from a thesis submitted by FrancoAjmar to the Committee on Genetics of the University of Chicago in partial fulfillment of therequirements for the degree of Doctor of Philosophy.

t Recipient of a grant under the Mutual Educational and Cultural Exchange Act and onstudy leave from the Istituto di Patologia Medica, University of Genoa, Italy.

I Research and Career Development Awardee of the U.S. Public Health Service.1 Carson, P. E., C. L. Flanagan, C. E. Ickes, and A. S. Alving, Science, 124, 484 (1956).2 Yoshida, A., these PROCEEDINGS, 57, 835 (1967).3 Carson, P. E., S. L. Schrier, and A. S. Alving, J. Lab. Clin. Med., 48, 794 (1956).4 Carson, P. E., S. L. Schrier, and R. W. Kellermeyer, Nature, 184, 1293 (1959).5Carson, P. E., W. K. Long, and C. E. Ickes, Federation Proc., 20, 64 (1961).6 Carson, P. E., G. T. Okita, H. Frischer, J. Hirasa, W. K. Long, and G. J. Brewer, in Pro-

ceedings of the Ninth Congress of European Societies of Hematology, Lisbon (Basel/New York:S. Karger, 1963), p. 655.

7 Carson, P. E., F. Ajmar, F. Hashimoto, and J. E. Bowman, Nature, 210, 813 (1966).8 Glock, G. E., and P. McLean, Biochem. J., 55, 400 (1953).9 Bowman, J. E., P. E. Carson, H. Frischer, and A. L. de Garay, Nature, 210, 811 (1966).

10 Dodge, J. T., C. Mitchell, and D. J. Hanahan, Arch. Biochern. Biophys., 100, 119 (1963).11 Zatman, L. J., N. 0. Kaplan, and S. P. Colowick, J. Biol. Chen., 200, 197 (1953).12 Sumner, J. B., Science, 100, 413 (1944).13 Ashwell, G., in Methods in Enzymology, ed. S. P. Colowick and N. 0. Kaplan (New York:

Academic Press, 1955), vol. 3, p. 87.14 Kellermeyer, R. W., P. E. Carson, S. L. Schrier, A. R. Tarlov, and A. S. Alving, J. Lab.

(lin. Med., 58, 715 (1961).15 Ben-Hayyhi, CI., A. Hoclunaii, and Al. Avroji, J. Biot. (hcmn., 242, 2837 (1967).

544 PROC. N. A. S.

Dow

nloa

ded

by g

uest

on

June

1, 2

020

VOL. 59, 1968 BIOCHEMISTRY: AJMAR ET AL. 545

16 Colowick, S. P., N. 0. Kaplan, and M. M. Ciotti, J. Biol. Chem., 191, 467 (1951).17 Gasiorowska, I., and K. Raczynska-Bojanowska, Bull. Acad. Pol. Sci., 11, 417 (1963).8 Alivisatos, S. G. A., and 0. F. Denstedt, Science, 114, 281 (1951).19Bergren, W. R., G. Donnell and W. Ng, personal communication.20 Kaplan, N. O., S. P. Colowick, and A. Nason, J. Biol. Chem., 191, 473 (1951).21 Kirkman, H. N., Nature, 184, 1291 (1959).22 Pontremoli, S., A. De Flora, E. Grazi, A. Mangiarotti, A. Bonsignore, and B. L. Horecker,

J. Biol. Chem., 236, 2975 (1961).23 Davidson, R. A., Ann. Human Genet., 30, 355 (1967).24 Kaufman, B. T., and N. 0. Kaplan, Biochim. Biophys. Acta, 39, 332 (1960).

Dow

nloa

ded

by g

uest

on

June

1, 2

020