The Isolation of Flavin Nucleotides1 - Journal of the American

Publication Date: November 1952. ACS Legacy Archive. Note: In lieu of an abstract, this is the article's first page. Click to increase image size Free...
0 downloads 0 Views 743KB Size
methyl iodide. A portion (84.3 g.) of the esters when fractionated yielded 3% methyl caproate, 26: (4 mm.), 26% methyl 2-phenylbuten-3-oiate, 84'( 2.7 m m . ) , ;?%of unidentified monomethyl esters, 88" (1.5 rnni.), 39% methyl pbenzalpropionate, 103" (1.9 mm.), 4% of a n unidentified fraction 110' (1.8 mm.), a 5% methyl ester of a dicarboxylic acid of composition approximately CuH1404, Tu. 125" (0.5 mm.), 5% (determined by calculation of saponification equivalents) and 6% of residue. The uniclentified monomethyl esters in fraction 3 gave some evidence that suggehted o-propenylbenzoic acid might be present but the i-olation of pure material proved more tlif3icult than the time available for the work permitted. Activity of the Methylstyrenes in Polymerization of Butadiene.-a-Methylstyrene, 0.25 mole, was rnet:ilated b y arnylsotiium prepared from 1 g. atom of sodiunl, 0.,52 mole of amy1 chloride anti 0.20 mole of isopropyl alcohol iri 500 ml. of pentane. This mixture was allowed to 5tand 9 days before its activity in polymerization of butadiene by bottle polymerization was tested. The conversions were 10, 30, 80 and 100% in about 60 minutes when 10, 20, 40 and 60 ml. of the reagent were used. The corresponding intrinsic viscosities of these four polymers were 0.40, 0.34, 0.38 antl 0.39. The corresponding gels were 17, 0, 19 and 17%. Three similar preparations of B-meth!lst)-ren?.lsodiuin reagents were made but the third wie gave typical results. The conversions of butadiene to a polymer were 14, 34, 80 aiitl 90% as 10, 20, 40 and 60 nil. of the reagent were used. T h e corresponding intrinsic osities were 1.5, 1.7, 0.9 antl 1.O and the corresponding gels were 26, 32, 28 and 43%. The effect of the methylstyrenes was tested also on an Alfiii catalyst of known activity prepared in the usual way (two separate preparations) from a total of 6 g. atoms of sodium, 3 moles of amyl chloride and 1.2 moles of isopropyl :ilcohol, antl gnseous propylene. The whole was diluted

[COUTRInTT I I O Y FROM TIIE I U S T I T U T E FOR

t o a volume of 5250 ml. and stored in a well stoppered bottle. This master batch, which should have a concentration of 0.23 milliequivalent of allylsodium per ml. was treated in 250-ml. aliquots with 0 , 1.69, 3.38, 6.67 and 13.5 g. of Qmethylstyrene, the last quantity being twice the amount needed to react with all of the allylsodium. Each of these five aliquot portions was stirred for one hour at 8000-9000 r.p.m. a t 25 f 5' and then transferred t o well stoppered bottles and diluted t o 350 ml. with dried n-pentane. Tests for the catalytic activity caused by 5 ml. of the reagent on 30 ml. of butadiene in 210 ml. of pentane were made a t the eiitl of i, 29 and 167 days. The control (or no. 1) caused fi4, :SI atid 41% of polymerization. The corresponding intrinsic viscosities of the polymers were 14, 11 and 11 and the gel contents were 84, 22 and 60. The fifth reagent, that is, the one with the most a-methylstyrene caused a little more polynierization, 66, 61 and 57%. The corresponding intrinsic viscosities were 16, 13 and 12 and the gels were 50, 43 and 47. The intermediate preparations gave conversions and other values within this general range. Actually, out of a total of twelve tests, the controls were exceeded seven times. The effect of O-methylstyrene was tested in the same way hut the catalysts were allowed to stand for 1 and 36 days before each test. The control (or No. 1) showed 45 and 35yo conversions. The intrinsic viscosities were 19 in 1,oth cases aiid the gels were 77 and 80%. The fifth sample, which contained the largest amount of P-methylstyrene, caused 57 and 39y0 polymerizations. The corresponding intrinsic viscosities were 18 and 16 and the gels were 50 a i d 42%. In every test (eight comparisons) the reagents with 0-methylstyrene showed a n increase over the corresponding control, and the usual increase was 20 to 30%. CAMRRIIXE 39, kIASS.

ENZYME RESEARCH, UNIVERSITY O F

kvISCONSIU]

The Isolation of Flavin Nucleotides' €317

E. DIMANT,? D. R. S A N A DAIN~D F. AI. HUENNEKENS~ RECEIVED MARCH17, 1952

Based upon chromatographic methods a procedure is described for the isolation of the coenzyme, flavin-adenine-dinucleotide (FAD), from animal tissues and yeast. The authenticity arid purity of the product are established by means of ( a ) enzymatic assay; (1)) absorption spectrum: antl (c) pauer chromatography i l l several solvent systems.

The yellow compound, flavin-adenine-dinucleotitle (FAD),j which functions as the prosthetic group for many oxidative enzymes, was first isolated and characterized by Warburg and Christian .6,i Their method consists of extracting the flavins from yeast or animal tissues with hot water and trichloroacetic acid and subsequently purifying the F-iD by means of repeated extractions into (1) A report of this work was presented earlier-D. R. Sanadi and 1:. &I. IIuennekens, 117th hleeting of t h e American C h r r n i r a l Society, Detroit, Mich., April, 1950,Abstracts, p. GOC. (2) Fellow of t h e S a t i o n a l H e a r t Foundation, U. S Pril)lic Health Service. Federal Security Agency.. ( 3 ) Senior Research Fellow of t h e National Cancer I n s t i t u t e , U S Public Health Service, Federal Security Agency. (4) Department of Biochemistry, IJriiversity of U'ashington. Seattle, lvashington. ( 5 ) T h e following abbreviations will be used throughout this paper: F.4D, flavin-adenine-dinucleotide; F M N , flavin mononucleotide; K b , riboflavin; CoA, coenzyme A; A M P , adenosine-5-phosphate; A D P , adenosine diphosphate; A T P , adenosine triphosphate; D P N , diphosphopyridine nucleotide; T P N , triphosphopyridine nucleotide. ( 6 ) 0.Warburg a n d W.Christian, Biochcm. Z., 898, 150 (1938). (7) For a n excellent review of t h e flavins a n d t h e flavin nucleotides, see H.Theorell, "Die Alloxazin-Proteide (gelhe Fermente)," in M e t h oden d e r Permentforschung, Vol. 111, edited by E. Bamann a n d K. Mgrback, moto offset Reproduction by Acadeu&icPress. Xnc.. New York. N. Y..1845, pp. 2361-2384.

phenolic solvents, selective precipitation as the silver, lead or mercury complexes, and, finally, recrystallization of the barium complex. The ultimate preparation of pure FAD by the above method is partially offset by the disadvantages of a low over-all yield and by the contamination with other adenine nucleotides until the final stages of purification. In an effort to overcome these difficulties a method has been devised which makes use, principally, of the following chromatographic techniques : (1) adsorption chromatography on Florisil which seDarates the extracted flavins from other nucleotides; and (2) partition chromatography between phenol-butanol and water on Celite, or adsorption Chromatography on dicalcium phosphate, which separates the various flavins. The FAD is obtained in a 4&5Oa/, over-all yield and has a dry weight purity of 0.40-0.60; the remaining impurity is innocuous in various enzmatic test systems. The quantitative data on yield, purity, R value, &c., given for the various procedures in the experimental section, are, unless otherwise specifi.ed, average values derived from many individual preparations,

Nov. 5 , 1952

THEISOLATION OF FLAVIN NUCLEOTIDES

In addition to the isolation of flavin nucleotides, the technique of Florisil chromatography has general applicability in effecting a separation of other biologically important nucleotides. Byproducts such as: CoA, DPN, T P N , AMP, ADP and ATP are obtained as partially purified concentrates by this method.

544 1

Preparations. A. Preparation of FAD. 1. Extraction of Flavins from the Tissue.-Fresh pig liver is packed in ice

and transported immediately to the laboratory. A 1-kg. portion is cut into srnall pieces, divested of connective tissue, and extracted in the following manner with trichloroacetic acid: I50 g. of the tissue and 450 ml. of cold deionized (or distilled) water are ground in a Waring Blendor for one minute. 70 ml. of 50% (1:l by weight) trichloroacetic acid is added and the blendorization coiitinued for l / 2 minute. The deproteinized mixture is centrifuged in the Experimental cold for 5 minutes at 2300 X g. The opalescent, yellow supernatant is removed by decantation and filtered through Materials.--All of the organic reagents used it1 this investigation were of a C.P. grade and were used without fur- coarse, fluted paper to remove fat.and debris. The inther purification. Merck and Co., Inc., C . P . reagent stability of flavins in acid solution and in the presence of phenol is recommended because of its freedom from colored light (cf. Part C ) makes it imperative that the extraction from the tissue and subsequent Florisil chromatography be impurities. carried out as rapidly as possible and with a minimum exRiboflavin was either the commercial Merck or Eastnian posure t o strong light. product; additional flavins in pure form were generously 2 . Chromatography on Florisil. -A glass chromatosupplied as follows: flavin mononucleotide (synthetic) by graphic tube 3.5 X 30 cm. is filled with Florisil to a height Hoffmann-La Roche, Inc.; L-lyxoflavin (synthetic) by Dr. Karl Folkers of Merck arid Co., Inc.; and FAD (iso- of about 20 cm., and topped with a plug of glass wool. The lated by the Warburg and Christian by the Sigma adsorbent is not tamped, since this greatly reduces the rate of flow. The column is prepared in advance arid stored in Chemical Co. Florisil, a magnesium silicate adsorbent, was obtained the cold room ( c n . 5 ' ) . Water is passed through the column as the 30/60 mesh grade from the Floridin Co., Warren, Pa. just prior to use followed by the filtered trichloroacetic Hyflo Supercel and Celite (No. 585) were products of the acid extract (ca. 2600 ml.). The flavins remain adsorbed Johns-Manville Co. The Celite was purified by heating at the top of the column. A constant-level separatory with a n excess of 5% HzO? on a steam-bath for three hours, funnel is useful for increasing the rate of flow and avoiding whereupon it was filterecl, washed with deionized water and frequent addition t o the column. After the extract has passed through, the column is washed, in turn, with 2 1. of dried a t 100" for two days. 2% acetic acid, 2 1. of deionized water and finally with 2 1. The unsuitability of several commercial preparations of dicalcium phosphate for our chromatographic purposes led of 0.570 pyridine. It should be noted that during the water wash a very pale yellow color (not a flavin) moves down the to the following special preparation: Solution A, 150 g. of column and appears in the effluent (cf. Part D ) . The inCaC13.2H?O ( C . P . ) dissolved in 750 ml. of water. Solution B, 240 g. of K?HPOd3H20 dissolved in 1000 ml. of water. tense yellow flavin band, however, diffuses slowly down the After both solutions have been heated to 70-75', A is added column under the influence of mater and 0.5% pyridine, and to B with stirring. The precipitate, in the form of a thick the pyridine wash should be terminated if the band reaches the bottom of the column. sludge, is allowed to stand for 20 minutes and then filtered Following the washing with 0.5% pyridine, the flavins are on a buchner funnel, washed with 1000 ml. of deionized water without allowing the cake to dry and crack during eluted from the column with approximately 500 ml. of 5% the washing, and dried at 100" for 24 hours. The hard, pyridine. The yellow flavin solution is adjusted to p H 8, dry adsorbent is pulverized in a mortar, and the particles extracted four times with 3 volumes of chloroform (the between 40 and 100 mesh mixed with a n equal weight of Hy- chloroform being discarded), filtered through a waterflo Supercel. The temperature of precipitation is critical, wetted paper, and finallv warmed under reduced pressure and if carried out a t a lower temperature, a weaker adsor- (water aspirator). If the flavin has been washed adebent, easily crushed by hand, is obtained. Although 24- quately on the column, the pH of the 570 pyridine eluate hour drying is usually satisfactory, the activity of the ad- will be 7 or higher, the R value 5 or less, and no emulsions will result when the solution is extracted with chloroform. sorbent may be increased by prolonged drying. Prior t o the next step in the preparation, the flavin solution Xanthiiie oxidase was isolated from cream by the method may be stored in the frozen state or lyophilized to yield of Corran, rt a1 . a The split D-amino acid oxitlase was prepared from pig 0.190 g. of a fluffy, light yellow powder having a dry weight purity of 0.10 and a n X value (ratio of light absorption a t kidney by a slight modification of the methods of Krebsg arid 260 and 450 m p ; q f , Iliscussion) of 4.0-7.0. The K value i\:arburg and Christian.6 Methods.--Absorption spectra were determined in the for the original trichloracetic acid extract is c-a. 120. 3. Partition Chromatography between Phenol-butanol Beckman spectrophotometer, model DU. All determinavarious flavins are next separated from tions were carried out in quartz cuvettes of 1 cm. thickness and Water.-The each other by means of partition chromatography. The upon solutions a t p H 7.0 unless otherwise specified. Concentrations of flavin solutions were determined from column is prepared on the previous day in the following light absorption measurements at 450 mcc. From the usual manner: 1800 g. of phenol and 800 ml. of n-butanol arc equation log Io11 = ecd, the molal extinction coefficient mixed by warming on a steam-bath and then cooled to room temperature. 120 g. of washed Celite is added, followed by ( e ) a t this wave length is assumed to be 11.3 X lO3cm.2/mole 40 ml. of water from a buret or pipet. During the latter for FAD,6 (cf. Discussion) (below PH 4 , E = 9.83 X 103). Assuming the following molecular weights: FAD (C27H18- two additions the solution is stirred vigorously. The soluN901;P2, 785.6), F M S (C,7HaA-409P, 456.4) and R b tion is next stirred intermittently for about one hour and (C17H20x406, 376.41, it follows that solutions containing then allowed t o stand for about three hours. As much as 10 pg./ml. of each of these compounds will exhibit, respec- possible of the clear supernatant is decanted and saved as the values of 0.144, 0.249 and 0.301. Purity developing solvent. The residual slurry is used to prepare a tively, log (10/1) column cn. 14 cm. in height in a tube ca. 3.5 X 19 cm. For of a sample, on a dry weight basis, was obtained by use of this size column no more than 10 mg. of total flavin should the above considerations. Ascending paper chromatograms with the indicated sol- be chromatographed; hence, the samplc from Part 2 (above) tems (cf. Table 11) were ruri on 40 X 40 cm. sheets should be chromatographed on two separate columns. of U'hatman S o . 1 paper, the flavins being detected by The desired rate of flow (1-3 ml./min.) may be achieved by means of their yellow-green fluorescence when observed pressure from a constant-level separator? funnel or, prefunder ultraviolet light from a "Mineralite" lamp.lo The erably, by compressed gas (Nz or COZ, but not 0 2 ) . After preparation, the column is stored by leaving several cm. of details of these, and other, solvent systems and the RF fluid over the adsorbent bed and sealing each end of the tube values for various other flavins are given elsewhere." with a cork stopper. ( 8 ) H. S. Corran, J. G. Dewan, A . H. Gordon and D . E. Green, Prior to the partition chromatography, the flavin solution Biochem. J . , 53, 1694 (1940). from the Florisil column is treated with ammonium sulfate, (9) H. A. Krebs, ibid., 29, 1620 (1935). as described below; or, if the solution had been lyophilized (10) C. E. Carter, THIS J O U R N A L , 72, 1460 (1950) previously, the powder is extracted with successive 10-ml. (11) F. M. Huennekens, D . R . Sanadi, A . I. Schepartz and E. portions of water by centrifugation in a clinical centrifuge. Dimant. ,bad., to be submitted. Three or four extractions suffice t o remove most of the color

E.DIMANT, D. I ; ) , approximately 200 ml.,treated with one-half the weight while the first migrates rapitllj- downward and is collected (100 g , ) of ammonium sulfate arid extracted successivcl>- in test-tubes in lO-nil. portioni. When this first band has with 10, 5 and 5 ml. of the developing solvent. The com- bceii complctely elutctl from t h e column, the developiiig bined, dark-orange phenolic extracts are centrifuged for 10 wlvrnt is changed t o 2570 pyridine which causes slow miminutes at top speed in a clinical centrifuge, and the upper gration tlownrvitrd antl elution of the top band. It has been IaJ-er, carefully excluding the interface region, drawn off observed that more active batches of dicalcium phosphate with a pipet. T h e column is allowed t o drain to incipient (prepared b y I)rolongctl heating a t higher temperatures, i.r., dryness and t h e flavin solution in the developing solvent a t 130° \rill separate the mixed flavins into thrce F X X , F.IU and FAD-X. This advantage is offset, ;idded dropwise with it pipet-care I)eing taken not to tlisturb the adsorbent bed. Under gravity or tank pressure hoire\.er, by the fact t h a t the I'AD must now be eluted with the flavin solution is passed into the coluinn, agaiii to the 3 -l(JYo pyridine while the FXD-X can bc eluted only with point of incipient dryness. Five nil. of thc tie\-eloping \ol- coiicentrated p>-ridine which causes extensive breakdotvvl~ vent is added and the column again brought t o n e a r - d r ~ - n c ~ s . into i t i inotionucleotitle. \\'hen the flavin zone has been drawn completely into the i ~ l The fractions of the first l ~ L i ~are column the developing solvent niay be added iii quantity. papcr chromatograpl Since the flavins are added t o thc column in a phenol-butanol and 2 usually colibutanol -acetic acid d u t i o n which is saturated with respect t o water. equilib- tain exeluiively F1\1 xtures of F M X arid rium is not immediately reestablished ivith the bulk of the FAD, while the bulk of flavin, F A D , is found ill the remainsolvent on the column, which is less than saturated with ing tubes. It should be noted that if the co1um11 has not \rater. This is reflected b y the behavior of the flavinr on tweii tainped properl)., the mixture of FIMS and F A D \rill the column: Initially, owing to the high coneeiitratioii of esteiitl over in:iii>- tubes, heiicc decreasing the yield of pure witer, the combined flavins move rapidly doivn the colurnii VAII. Thc F A I ) fraction. are pooled and lyophilizctl to anti rnay appear to be resolving into individual bands. dryness !.ieltiiiig a light yellow powder weighing 115 g . antl L'pon reaching a point about one-third thc (list hxving i i dry weight purity of 0.10 (contaminated with clithe column, however, the diffusc yellow band ' calciuin phosphate) and :in K Lalue of 4,(1--4.:. T h e repped as the true equilibrium under the defii covery is isyo for this step. tem is established. At this point, ii l)ro\vii->5 . Further Purification of FAD.--The impuritiea resrates from the combined zoiie aiid, after its slow paswge maiiiiiig Lifter the chroiiiatogral)hic steps appear t o have been tiown the remainder of the colunin, is collected in the effluent derived both from the original tissue and froin the columns. 1 LU. 30-60 ml.). This flavin, FAD-X, will be described i n a Further purificatioti (up t o 0.50-0.60) rnay be achieved b>. later publication." It may be transferred to water anti repetitivc extraction into phenol, but this advantage ib irorked u p by the same procedure.; rvhich xi11 be described partially offset by the relatively high 105seS. The drjl~clowfor FAD. inaterial (10--20ing.) is dissolved iii 50 rnl. of water and amAfter elution of F A D S ib coiiiplete, the top of the colutnii, iiionium sulfate ( 5 g. per 10 tnl. of solution) added. Follo\vto a depth of 2-3 cm., is removed as a slurry by ineatis of a ing the procedure for the phenolic extraction in Part 3 , thc pipet. This mixture, which contains the sinal1 amount of solution is extracted three times with 2-mi. portions of ;I I:MN present in the original flavins, may be discarded. niisture of 10~)g. of phenol and 50 ml. of benzyl alcol~oi. The developing solvent is now changed by adding one part The combined phenol extract is dried over freshly ignited of water to each 30 parts of original solvent, and the chro- ,odium sulfate, decantcti, filtered through a paper wetted matography continued with the wa tcr-enriched mixture. n i t h phenol-benzyl alcohol, and the sodium sulfate washed T h e F A D moves slowly down the column as a discrete with two 5-ml. portions of previously dried, phenol-benzyl >.ellow zone and is collected in the effluent ( c a . 100-300 ml.). ,ticoho1 mixture. The flavins are transferred from the I n a large separatory funnel the effluent is mixed with :3 ~ihenolicsolution to water in the manner described prcvolumes of chloroform and extracted with 25 ml. of water. viously (cf. Part 3 ) aiid lyophilized t o dryness. 'i'lic organic layer, in turn, is re-extracted with t m or niore Precipitating FAD a t pH 2 :ts t h e lead or silver complex ;ctiditional 25-1111. portions of water until it is colorlesP, aiid : t i i d rccovering the flavin b y treatment with H2S6 does not the conibined water fraction containing the flaviiii i\ eslcari t o :tn>- iigliificarit incrca5e in purity. Furtherinow, t r x t e d three times with 200-1111. portions of chloroform, t h e c \ - c n under r~iiltlc o ~ ~ t l i t ithcre o ~ ~ sis : t l \ r a y s sonic breakdo\rrn chloroforrn being discarded subsequently. The flavin solu- of F A D to F M S . tion is freed from traces of chloroform by the procedure ticT h c relatively high solubility of the bariurrl salt of F.11) kc~ribedpreviously under Part 2 , ;ind lyophilized to dryness i n water 12.5 nig. 'iiil. at nt5")6 does not permit purification The product (original flavin froin the Florisil column di- by recrystallization unless largc quantities are available. \-itled, chromatographed on two separate Celite colurmii, I€owever, F A D is conveniently precipitated from solution ~ i i dthe F A D fractions recombined) is an oraiigc powder a t pH 8.7 b y arldiiig barium acetate and 5 volumes of cold \\-eighing 38 mg., having a dry weight purity of n.30 and an acetone. \ \ h i 1 decomposed with a n excess of sodium sulI? value of 3.8-4.0. Thc recover>- of F.4D is 55-iflVc iii fate the recovery of F A D is quantitative with no breakdown t his step. to 1;MS. Precipitation b y acetone in the ahseiicc of barium 4 . Adsorption Chromatography on Dicalcium Phosphate. lends to extensive breakdown of F A D . - --.4dsorption chromatography on dicalcium phosphate B. Use of Other Tissues.-Both heart and kidney may offers an alternate method which can be used in place of iilso be u w l as a source of F.4D although the yields are partition chromatography for separating the flavins. T h e oinewhat lower than in the case of liver. Tissues may be dry adsorbent is carefully packcti in a glass chromatographic tored in the frozen state prior to use without decreasiiig tube (1.3 X 15 c m . ) t o a height of about 10 em. Thc ca- the yield of F A D ilacity of this column is about 20 mg. of flavin. Thc ariBaker,' ycasi (Ked Star) may also serve as the starting sorbent should be added der gentle suction from a water material, although in this case the usual procedure for esratc portions, e:tch one hcing traction of the flavin is modified as follows12: 3 kg. of yeast aspirator) in about 10 cvenly tumped. i, crurnbletl and tatddctl to 6000 cc. of water kept a t 70-80"; Before adding the flavins, water is allo~vedto pa\s through the solution is kept a t this temperature for 10 minutes after the column aiid it is observed t h a t the rate of travel through the lait of the yeait has been added and then rapidly cooled lhc adsorbent itself appears t o differ from that between the to room temperature bv the addition of ice or by means of a adsorbent arid the glass wall. The water aspirator is ad- cooling coil. The solution is centrifuged for 20 minutes a t justcd so that by visual observation these two solvent fronts 2300 X g, and the supernatant is filtered through a pad of nearly coincide. Celite 1 cm. in depth. Solid trichloracetic acid is added t o 'The mixed flavin sample, previously lyophilized after the the filtrate t o make a final concentration of 5 % , the solution 1;lorisil chromatography, is extracted by centrifugation centrifuged for 10 minutes a t 2300 X g, and the supernawith successive 3-ml. portions of water until the extracts t a n t processed thereafter in the same manner as the acid ex:ire essentially colorless. The combined estracts ( c a . 10-15 tract-from liver. rill.) are added with a pipet t o the top of the columii, which The yield I)F i:.A I 1 Iron1 vxiuus animal tissues and froin licih been brought to the point of incipient dryness. After the flavins have been conipletely brought into the column, i w t e r is added anti thc chroin;Ltography continued. Thr.

THEISOLATION OF FLAVIN NUCLEOTIDES

Nov. 5 , 1952

bakers’ yeast is shown in Table I. I t should be noted that these figures are values after the Florisil chromatography and represent only one extraction from the tissue.13 The dry u-eight purity of the flavins at this stage is ca. 0.10 in all cases except for yeast which is ca. 0.03 (the R value is likewise high in the latter case, ;.e., 10-20).

TABLE I YIELDSOF F A D

FROM

Source

Pig liver Pig kidney Pig heart Rabbit liver Pigeon breast muscle Bakers’ yeast

VARIOUSSOURCES Flavin (as F A D ) , rng./kg. of wet tissue

19.0 12.6 7.5 11.0 9.4

18.5

C. Stability of Flavin Nucleotides.-The instability of F A D below pH 3 and above pH 10, a t elevated temperatures and in the prescnce of light has been confirmed using paper chromatography (phosphate buffer system). Unless the paper chromatograms are prepared and run in near darkness, a spot of lumichrome14 ( R F = 0.05 in the phosphate system) will always be observed, even with pure samples. Although in the first stage of purification F A D is chromatographed on a Florisil column without breakdown, more purified samples undergo extensive decomposition t o F M N if rechromatographed under the same conditions. This anomaly, noted also by Bessey, et UE.,’~ would appear to be due either to a protective action of impurities in the crude state or subtle differences in experimental conditions during the first chromatography. D. Nature of the Various Effluents from the Florisil Column.-Since many nucleotides, other than the flavins, are extracted from the tissues by trichloroacetic acid, it is of interest to ascertain their location in the various eluates from the Florisil column. The trichloroacetic acid and acetic acid effluents from the column contain no nucleotides according t o the criterion of paper chromatography. Adenine-containing nucleotides are found in high concentrations in both the water and 0.5% pyridine effluents, and niay be detected by their characteristic dark blue “quenching” spots on paper chromatograms using 5y0 iXa2HP04 as the developing solvent.1° If the water wash of the column is increased to 3 l., essentially all of the CoA is found in the pale, straw-yellow effluent.l6 As yet no attempt has been made t o isolate the above nucleotides, but the ability of the Florisil column to effect a preliminary separation of these compounds from a crude extract suggests the possibility t h a t this method might be used to prepare “concentrates.” After the flavins have been eluted with 5y0 pyridine, a pale yellow band is often observed at the top of the Florisil column. Eluted with 25% pyridine, this material is identified as FAD, but no explanation can be offered for this apparent anomaly where a small fraction (ca. 2 4 % ) of the total flavin is tightly bound t o the top of the Florisil column.

5443

(bj d second criterion of purity is provided by the method of paper chromatography using several different solvent systems (cf. Table 11). The phosphate buffer system is simple and extremely useful, for it achieves a clean separation in the case of riboflavin, F M N and FAD; the only disadvantages are the diffuse nature of the spots and the slight break‘down of the various flavins to lumichrome (RF = 0.05). The formation of this degradation product may be minimized by preparing and running the chromatogram in the absence of light. A second system, composed of butanolacetic acid-water, has the disadvantage of relatively low RF values for the phosphorylated compounds, but the advantage of small, discrete spots which permit of the assignment of accurate RF values. The third solvent system, phenolbutanol-water, effects the separation of FAD and FAD-X. RF

\

T

OF THE ~

TABLE I1 FLAVISS ~ BY ~ ASCENDXSG ~ ~ PAPER CHROMATOGRAPHY

Solvent Systems: 1, 5% SapHPOa in waterlo; 2 , 4/1/5 nbutanol/acetic acidlwater (top phase)’8; 3, 160 g. phenol/30 ml. n-butanol/lOO ml. water (lower phase). 7 -

Flavin

FAD FAD-X

FMN Rb

Solvent system

1

2

3

0.35 .35 .48 .30

0.05 .05 .13 .30

0.23 .47

.I7 .80

Using these systems, the purity of the FAD preparation may be verified a t each step. No spot is detected a t the expected positions of riboflavin and only occasionally a faint spot a t the position of FMN. This test for the presence of breakdown products of FAD is rendered even more delicate because of the 7-fold greater molal fluorescence of riboflavin and F M N as compared to FAD.14 After separation of the flavins on Celite or dicalcium phosphate, the three components FAD, FAD-X and FMN, each exhibit single fluorescent spots in each of the solvent systems. (c) The absorption spectrum of F 4 D (purity 0.5) is shown in Fig. 1. I t is apparent that the ratio, R,of light absorption a t 260 and 450 mp, may be used as a criterion of purity for a flavin Discussion nucleotide. Contaminants, such as other adenineThe authenticity and purity of the FAD have containing nucleotides or traces of organic solvents been established by the following criteria: (a) (ie., phenol and pyridine), will absorb at 260 mp enzymatic assay in the D-amino acid oxidase sys- but not a t 450 mp, and the R value will be increased tem, (b) paper chromatography in various solvents over that of the pure flavin. It is of some importance, therefore, to establish the correct R value for and (c) absorption spectrum. (a) In the D-aminO acid oxidase s y ~ t e m , ~pure , ~ FAD. The commercial availability and known molecthe FAD is fully active a t each stage of preparation, ular weights of pure R b and FMX make i t possible the impurities being innocuous. l 7 to determine the absolute values of the molecular (13) Under t h e conditions described in P a r t 1 , a single extraction of extinction coefficients, E , for these substances. t h e tissue with trichloroacetic acid liberates 80-85% of t h e total flavin. Since the preparation of pure FAD is not as readily (14) P. Karrer a n d T.Korber, Helo. Chim. Acta, 18, 266 (1935). accomplished, the R value, being independent of (15) 0. Bessey, 0.H . Lowry a n d R. H. Love, J. Bid. Chem., 180, 756 (1949). the dry-weight purity, may be used, nevertheless, (16) F. M . Huennekens, H. R. Mahler a n d E. D i m a n t , unpublished as a measure of the purity relative to other nucleoobservations. (17) We are indebted t o Dr. H e n r y R. Mahler f o r assaying our preparation in this system. Using Protein E of Negelein a n d Bromel (Biochem. Z.,300, 225 (1939)) from pig kidney as a source of t h e enzyme, a value of t h e Michaelis constant, K,, equal t o 3.3 X 10-7 M

was obtained This may be compared t o t h e value 2.5 X 10-7 M reported by Warburg a n d Christianfir’ for their F A D preparation tested with t h e corresponding sheep kidney enzyme. (18) J. L. C r a m m e r , Nature, 161, 349 (1948).

IC. DIMANT, L). K. SANAJX AND E'. M. HUENNEKENS

5444

I-

-.

VOl. 74

-

however, Ballz2has obtained a value of 3.3-for FAD released from xanthine oxidase. In principle this 1 method is capable of giving accurate values for FAD, since purification of the enzyme should climinate all contaminant nucleotides and leave only the bound prosthetic group. The possible existence of a second prosthetic group in the case of xanthine oxidase and aldehyde oxidase renders the data in these cases less reliable but the Lamino acid oxidase data would seem to be unI5 equivocal since FAD is the only known prosthetic group of this enzyme. Furthermore, i t is unlikely that remnants of native protein could survive the purification procedures in sufficient amount to 250 300 350 400 410 5 34 interfere with the R determination. A h mp). Finally, the R value of FAD may be calculated Iiig. l.--Absorptioii spcctruni of PAL). as 3.f.iZ3on the assumption that the light absorption in the dinucleotide is additive for the isoalloxazine tides. Thc E values in Fig. 1 are calcukated 011 the moieties. Warburg and Christian6 considered that basis that e = 11.3 X l o 3a t 450 nip.6 their lower experimental value could be explained From Fig. 1, R is determined to be 3.S and no on the grounds that the absorptions were not sample of FAD was encountered in this investiga- additive and this position is apparently strengthtion where R was lower, unless breakdown to FMN ened by the finding14that a t pH 7.0 the fluorescence had occurred. From the data of Warburg and of FAD is only 0.14 of that for FMN or Rb, but Christian,6however, R is calculated as 3.3 and it is that as the pH is lowered, the fluorescence of FAD possible that our best preparations (purity 0.50') rises until a t pH 2 it is equivalent to that of the still are contaminated by relatively large amounts of mononucleotide. -4 possible interpretation of these material which absorb a t 260 nip, I n support of fluorescence data, suggested by Bessey, et the higher value, however, are the findings of several is that in FAD the adenine acts as an "internal investigators who have reported spectral data on quencher" of the isoalloxazine nucleus, and that a t FAD obtained by deproteinizing purified flavo- low pH values, this interaction is prevented, proteins containing this nucleotide as the prosthetic perhaps owing to a change in configuration. While group, and subsequent purification of the flavin. this explanation is probably correct, it does not and Gordon, et aZ.,'$found R follow from any theoretical considerations that such Thus, Corran, et valueszoof 3 . i for FAD liberated from milk xanthine an interaction should affect the absorption of light oxidase and liver aldehyde oxidase, while from the by the individual absorbing structures. data of Singer and KearneyZ1i t may be calculated Acknowledgments.-The authors are indebted to that R = 4.0 in the case of the pure L-amino acid Professor David E. Green for his interest and enoxidase from snake venom. Recently we have couragement in this problem. I t is likewise a been informed by Dr. Henry R. Mahler (private pleasure to acknowledge the helpful discussions communication) that his preparation of highly given on various aspects of the problem by Drs. purified DPN-cytochrome c reductase from beef I-I. R. Mahler, 11. 1'. I3uel1, A I. Schepartz and heart contains a flavin prosthetic group, which, G.C. Xueller when resolved from the apoenzyme, has an R value of 3.9. Contrary to the above findings, \IEATTL.E, b~'ASH1AILIOZ "-

.

(19) A. 13. Gordon, D. E. Green a n d V. Subrahmanyan. Biochriit. J . , 34, 764 (1939). (20) I n the t w o cases cited, the R value is based upon readings a t 265 mp rather than 2 6 0 m p . As seen in Fig. 1, however, the change in absorption a t 260 a n d 265 mp is only 2-3%',. (21) T. P. Singer a n d E. B. Kearney, Arch. Biochein. 27, 348 (1950).

( 9 2 , E C . Ball. J . B i d . Chem., 128, 5 1 (1939). ( 2 . 3 ) Dr. C. S \Vhitby has informed us t h a t pure riboflavin has e = I ? L' X 10: at $SO mp, in agreement with Singer a n d Kearney.*' i l t i r i e = 28.0 X 1 0 . a t 2G0 m p . Assuming t h e same values for F51S ani1 from t h c known values (H. Kalckar, i b i d . , 167, 445 (1949)) of c = 15 !I X 103 f o r A M P , the calculated R value for F A D , assuming 15.9)/12.2 = 3.6. aciditivity, would be (28.0

+