Enzymatic Reduction of Nicotinamide Adenine Dinucleotide Phosphate Induced by Radiolysis*

Radiolysis in the presence of ferredoxin-NADP reductase converts NADPf to NADPH; without the enzyme, a biologically inactive product, presumably the reduced dimer (NADP)*, is formed. The enzyme specificity on radiolytic reduction resembles that with more conventional reductants. The irradiation energy is supplied as x-rays or Z-m.e.v. electrons to a phosphate-buffered solution in the absence of oxygen with ethanol or methanol to scavenge oxidizing radicals.

reductase converts NADPf to NADPH; without the enzyme, a biologically inactive product, presumably the reduced dimer (NADP)*, is formed. The enzyme specificity on radiolytic reduction resembles that with more conventional reductants. The irradiation energy is supplied as x-rays or Z-m.e.v. electrons to a phosphate-buffered solution in the absence of oxygen with ethanol or methanol to scavenge oxidizing radicals.
l'yridine nuclcotitlc oxidation and reduction in biological systams is frequently catalyzed by flavin enzymes. A 2-oq reduction of the pyridinc ring and the addition of a proton results in the formation of a higher energy compound (1). The mechanism of this process has bcrn the subject of earlier investigations in this laboratory.
I'hotoreduction of the nicotinamide adenine tlinucleotide SAI>l'+ was show-l1 to occur in the presence of cithcr of two synthetic pigments structurally unrelated to chlorophyll (2) and t,hc flavoprotem ferrcdoxinNAD1' rcductase (2). The formation of a biologically active rcduccd product was found to br dependent on the presence of the enzyme.
This finding suggested to us the possibility that radiolysis could be used to produce a similar reduction of Nhl)l'+ to the biologically active product NADI'H.
The irradiation of neutral water by x-rays, y-rays, or high energy clcctrons gcncrates the following species within 1OW s from the time of energy absorption (3)  viewed the studies on pyridine derivatives including the pyricline nucleot.ides.
Radiolysis in the absence of enzymes produces a single equivalent reduction of the pyridine ring via a radical mechanism (6, 7) ; two of the radicals formed then react to yield a stable species thought to be a reduced dimer.
The final product is inactive with the lactate and ethanol substrate dehydrogenases (6) and wit,h spinach ferredosinNAD1' reductase flavoprotein (this paper).
The present paper reports the production of NADPH during steady state irradiation of NADP+-containing solutions in the prcsencc of a flavoprotein with ethanol or methanol to act as scavengers for hydroxyl radicals (10). The presence of the flavoprotein, ferredoxinNAD1' reductasc, is required for the formation of NADPH; in the absence of enzyme, the biologically inactive dimer (5-7) is produced.

METHOD
Radiolysis is performed in a phosphate-buffered aqueous solution containing 0.1 to 1 mM NADP+ and 0.1 to 1 M ethanol or methanol. The flavoprotein is added as needed in 50 rnM phosphate buffer, pH 7.4. Solutions are irradiated in rectangular quartz cells (30 X 10 X 5 mm). Oxygen is removed by evacuation or by bubbling argon through the solutions and subsequent measurements are made under anaerobic conditions.
The irradiation employs either a Philips Electronics x-ray tube operating at 75 kv or a Van de Graaff accelerator operating at 2 m.e.v. The dose rates are, respectively, 170 or 6000 rads per s; the dosage is measured with an air-saturated solution of 10 rnM FeSOa in 0.8 N H,SOd (11,12).
Optical spectra are measured in a Cary model 14 spectrophotometer before and after irradiation; the fluorescence spectra in a Hitachi Perkin-Elmer MPF-2A spectrometer. NADP+ and NAD+ are used as purchased from Sigma Chemical Co. Ferredoxin-NADP reductase is purified from spinach according to Forti and Sturani (13) and stored at 20 PM in 50 mM phosphate buffer, pH 7.4.

Dose Rate and Enzyme Concentration--The
irradiation of an aqueous solution of iYADP+ in the presence of ethanol or methanol plus the flavoprotein ferredoxinNAD1' reductase yields two reduced products whose relative amounts depend upon the enzyme concentration and the dose rate. In the absence of the enzyme only the dimer, referred to as (NADP),, appears. It is presumably formed from two pyridine radicals as suggested by Land and Swallow (7). With the reductase enzyme present at 50 to 100 nM, x-ray irradiation at 170 rads per s produces only The (NADP)2 is not converted to NADPH by treatment with the reductase after irradiation.
The products of radiolysis have been characterized by the following five properties.
1. Optical spectra of the enzymatic and nonenzymatic products are shown in Fig. 2. The spectrum of the former coincides with that of NADPH.
The nonenzymatic product (NAI)1')2 has a broader absorption band in the 340-nm region and an cxtinction coefficient approximately 157, less than the value for NADPII, i.e. E:: nm = 6.22 X lop3 M-I cm-' (14). 2. Fluorescence excitation of the radiolysis products in the 340.nm absorption region gives different results in the presence and absence of enzyme. The enzymatic product has the 460 nm emission characteristic of NADPH in both intensity and position whereas the nonenzymatic product presumed t.o be (NAD1')2 has an emission maximum at 445 nm with only onetenth the intensity of the enzymatic product.
3. Stability in solution is less for (NADP)2 than for NADPH. Their respective lifetimes in the irradiated solutions are about 2 and 18 hours. Neither is particularly sensitive to OZ and both are acid-labile.
On the addition of HCl or by bubbling CO2 through the solution, both undergo rapid decomposition.
The acid sensitivity caused us to select conditions which would hold the pH after irradiation to the region between 6.8 and 7.6. In the dilute buffer 0.25 mM phosphate, pH 7.6, a drop of about 0.5 pH unit occurred during the irradiabions.
As shown in Fig.  1, this change was not important in t.hc determination as indicated by the identical data for NADPH accumulation in 10 mM phosphate buffer, pH 7.6, in which the pH remained constant.
4. Enzymatic oxidation of reduced products from radiolysis by frrricyanide followed by decrease in absorption at 340 nm is shown in Fig. 3. Fig. 4 illustrates the effect of the reductase concentration during irradiation on the fraction of the NADP+ converted to the enzymatically active NADPH.
(NADP)s dimer can be oxidized by ferricyanide but the rate is enzymeindependent.
With equal equivalents of (NADP), and ferricyanide the oxidation is completed in about 15 min. The prod- The biological reduction of S.WP+ to XADl'H by the Bavoprotein ferrctlosin~~4Dl reductase requires two electrons and a proton.
The flavin prosthetic group of the enzyme has three documented oxidation states: the oxidized, the one-clcctron reduced semiquinone, and the two-clcctron, or fully reduced (5,15,16). The flavoprotcin can extract electrons from one-clectron donors and transfer them to two-electron acceptors, such as NADl'+.
The way in which the rlectrons are transferred, in particular the roles of one-and two-electron reduced states of the enzyme in biological osidoreduction, is still unclear. Our experiments show that, the reduction of NADl'+ to the biologically active NADI'H can bc initiated by ionizing radiation coupled with the tlavin enzyme ferredosinXAD1' reductasc. In the absence of the cnzymt a biologically inactive reduced compound, probably the dimrr (XA41)l')z, is formed.
Land and Slvallow have given rate constants for the production of (X.4D)Z via hydrated electrons (7)  uct of radiolgsis reduction in the presence of the flavoprotcin undergoes a catalyzed oxidation rate equal to that of NADPH, indicating this product to be XADI'H. 5. The amount of P;XDP+ reduced in the presence and the absence of enzyme is approximately equal. For example, starting with the 100 PM iXADl'+, 17 krads of irradiation form approsimatcly 35 PM reduced XADl'H or (NX1>l')2 as indicated in Table I. If the latter is indeed a dimer, thr amount' of residual, nonreduced NAl)P+ should be different in the two cases. Therefore the two solutions were assayed for residual NADl'+ by the addition of an excess of glucose 6-phosphate and its dchydrogenase and the total free oxidized pyridinc nucleotide measured by t,hc increase in absorption at 340 nm. The concentration in the espcrimcnt containing the enzyme during irradiation showed a residual NAl> I'+ of 56 PM, whereas the enzyme-free: sample con tainrd only 33 phr. These data support the contention that the irradiation product in t.hc absence of enzyme is in fact a dimcr. Although the reconversions are not quantitative, 103y0 in the presence of the dimcr and 91%; in the presence of the enzyme, when one accounts for the difference in extinction corfficients in these experiments, the data arc r\ithiri the cxperimcntal (,rror. Enzyme and Nucleotide Specificity-The Ilavoprotcin ferre-dosinNAD1' rcductase is necessary for the formation of the biologically act,ivc reductant, NADPH, as was shown by substituting the heat-denatured for the native enzyme and by attempting to substikrte fret F,41> for the enzyme. Flavoprotcin n hich had been allowed to stand at room temperature for 48 hours in the presence of air lost both the ability to catalyze tht: oxidation of SAT~1'I-I in the presence of ferricyanide aud the ability to yield the cnzymatically active product during radiolysis. Fret FAD at concentrations of 50 rihi to 100 phf was similarly ineffective. The specificity for XAD+ versus X.41)1'+ of the flavoprotcin during irradiation was examined using the irradiation conditions reported in Table I and Fig. 4. The 30 IBM concentration of enzyme found effective for NADPH did not result in the formation of XADI-I. Wlwl the enzyme concentration was raised to 1 pM, as opposed to 30 11x, an appreciable yield of NADH was observed. Thus the selectivity of the Ilavoprotcin for the specific nucleotidc is the same in irradiation reduction as in reactions 1% ith the more conventional reducing agents (16). Almost identical results were found when X\jAI)+ was replaced by XADl'+.
In our experiments, the number of hydrated electrons reacting with enzyme molecules is fewer than the number reacting with NADI'+ molecules because of the high ratio of XAl>l'+ to fcrredosin-XADl reductase and the large rate constant of the ~AI_)I'.-l'roducillg reaction. I)irect reduction of the enzyme by either hydrated electrons or alcohol is unlikrly under our working conditions in view of the rate constants tlcmandrd at the reactant conccntrat,ion used. The direct reduction rate by hydrated clcctrons would have to be k = 1Ol5 ti1-i s-l, several orders of magntudr beyond diffusion limits (17). Direct reduction by alcohol radicals is similarly ruled out in view of their high reaction rate, k N 10" h1-l s-l (7) with NADI'+.
(NAl>l')B is formed by the collision of tno ?;ADI'. ; iVADPH can be produced in appreciable quantity only if the enzyme intercepts SADl' at a rate faster than the binary collision rate. The rate of dimcrization is given in our steady state case by the total amount, of dimer formed divided by the time of irradiation.
Vsing Table I and Fig. 4, a simple calculation shows that an enzyme molecule must shuttle an average of 30 reducing cq per s in order to produce NADI'H at a rate equal to the dimcrization rate. 'l'his number is an order of magnitude slower than the turnover value mcasurcd for the enzyme-catalyzed oxidation of NADPH by ferricyanide of about 300 reducing cq per s at room tcmperature (h_25O).
The most direct interpretation of the enzymatic interception of dimerization is a disproportionation of the pyridine radicals. The nature of the cnzymc-substrate complexes formed during this reaction cannot. be clarified at present; more refined techniques are ncrdcd.
The possibility of binding singly reduced NADl' " to enzyme, however, or more generally the occurrence of a singly reduced bound state, immediately presents itself.