This work has been digitalized and published in 2013 by Verlag Zeitschrift für Naturforschung in cooperation with the Max Planck Society for the Advancement of Science under a Creative Commons Attribution4.0 International License.

Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschungin Zusammenarbeit mit der Max-Planck-Gesellschaft zur Förderung derWissenschaften e.V. digitalisiert und unter folgender Lizenz veröffentlicht:Creative Commons Namensnennung 4.0 Lizenz.

FLAVIN-PHOSPHINE ADDUCTS 1023

On the Interaction of Flavins with Phosphine-derivatives

FRANZ MÜLLER

Department of Biochemistry, Agricultural University, Wageningen, The Netherlands

(Z. Naturforsch. 27 b, 1023—1026 [1972] ; received May 10, 1972)

Flavin-phosphine adducts, kinetics, light absorption, structure

The interaction of flavinium salts with triphenylphosphine is described. The complexes which can be isolated in high yields in the crystalline state contain one mole triphenylphosphine per mole flavin. Triphenylphosphine is covalently bound to the N (5)-atom of the isoalloxazine ring system. Dissociation constants and rate constants have been determined spectrophotometrically. Evidence for the structure of the complex was obtained from light absorption, infrared and nuclear magnetic resonance studies. The similarities between the flavin-triphenylphosphine adduct and the earlier described flavin-sulfite adduct are described and the possible biological relevance of these com-plexes is discussed.

The biochemists still depend largely on model studies for the postulation of structures of the inter-mediates observed with flavoprotein dependent reac-tions 1 . Therefore, recently many scientists have be-come active in the field of flavin model studies and it has become evident from such studies that the flavin molecule in the oxidized state undergoes a variety of reactions 2. However, many of these reac-tions are light catalyzed and, therefore, can not be directly correlated to the flavoprotein dependent biological reactions since most of them are dark reactions.

Recently, we described the interaction of free and some protein-bound flavins with a common sub-strate, namely sulfite 3~6 . These studies revealed the interesting fact that the flavoprotein oxidases are capable to form a sulfite addition product, whereas the flavoprotein dehydrogenases do not interact with sulfite 3 ' 4. In this context, the studies described below were undertaken to search for the possible interaction of flavins with other nucleophilic reagents.

Materials and Methods

The syntheses of the flavinium salts employed in this study were described elsewhere5. The solvents and chemicals used were of analytical grade. The light absorption studies have been performed either on a Cary 16 or on a Durrum prism-grating recording spectrometer, both equipped with thermostated cell holders. Cells of 1 cm light path have been used.

Requests for reprints should be sent to Dr. F. MÜLLER, Laboratorium voor Biochemie, Landbouwhogeschool, De Dreijen 11, Wageningen (The Netherlands).

The experiments were conducted at 25 °C. The dis-sociation constants and the rate constants were deter-mined using acetonitrile solutions of flavins and tri-phenylphosphine. The experiments were monitored at the long wavelength absorption maximum of the fla-vinium salt.

A typical example of the synthesis of the flavin-tri-phenylphosphine adduct is as follows: Triphenylphos-phine (2.0 g) was dissolved in abs. ether (30 ml). The slightly turbid solution was filtered and the flavinium salt (0.5 g) dissolved in acetonitrile (5 ml) slowly added with stirring to the filtrate. The solution turned red and got then gradually bleached. Stirring was con-tinued for 30 min at room temperature after which time the reaction mixture was allowed to stand at 4 °C for 10 hours. The slightly yellowish crystals were fil-tered and washed extensively with ether. In some cases it was necessary to add ether to initiate the crystalli-zation because a larger volume of acetonitrile had to be used to dissolve the flavinium salt. The yields were 60-95%.

Results and Discussion

In context with the sulfite studies we were interested to test the reactivity of flavins towards other nucleophiles. We have found that acetonitrile solutions of flavinium salts are bleached in the pre-sence of phosphine derivatives. In the described study we have used triphenylphosphine because of its air stability and easy handling. The bleaching reaction has many features in common with the flavinsulfite interaction5. Thus, the bleaching reac-tion occurs in the absence of light. Furthermore, the reaction is not influenced either by the presence nor absence of molecular oxygen. The product is stable towards molecular oxygen which proofs that the final product of the reaction is not the N ( 5 ) -

1024 F. MÜLLER

unsubstituted 1,5-dihydroflavin derivative as might be concluded from the light absorption spectrum of the product (Fig. 1) because N (5)-unsubstituted 1,5-dihydroflavins are extremely sensitive towards molecular oxygen. Moreover, the flavinphosphine

1.4

1.2 Q) <J c m -S

I 1.0

0.8

0.6

0.4

0.2

Fig. 1. Spectral course of dissociation of crystalline triphenyl-phosphine complex of 7,8-dimethyl-l,10-ethylene-isoalloxazi-nium Perchlorate upon dissolution ( 4 x l 0 — 5 M) in acetonitrile, at 25 °C. • • • about 2 min after the preparation of the solu-tion; 16 min later; 33 min later; — 59

min later; — • — • • 100 min later.

complex formation is quantitatively reversible leading to its parent compounds, i. e. flavinium salt and triphenylphosphine. This is demonstrated in Fig. 1. The spectral course of this reaction shows isosbestic points at 320, 287 and 248 nm. However, when more concentrated solutions ( > 1 0 - 2 M ) of the complex were prepared and allowed to stand at room temperature a deep red colored intermediate was formed rather slowly. After several days, the solution was colored yellow. The light absorption spectrum of the reaction mixture showed that be-sides the oxidized flavin also triphenylphosphine-oxide had been formed. This indicates that the flavin-phosphine adduct solution had been oxidized to the flavoquinonium level 7 and that the phosphine molecule was still covalently attached to the flavin.

The subsequent hydrolysis reaction yields then the uncomplexed flavin and triphenylphosphineoxide. No experiments under anaerobiosis have been done yet. This reaction is now under further investigation. This reaction has not been observed with the flavin-sulfite adduct.

The light absorption spectrum of a solution of crystalline flavin-adduct is given in Fig. I . In con-trast to the flavin-sulfite addition product which shows a well defined light absorption spectrum with a maximum in the 310 — 330 nm region 5 , the flavin-phosphine adduct exhibits a rather diffused light absorption spectrum extending beyond 450 nm with shoulders at 328, 298 and 248 nm. In fact, this spectrum resembles very much that of the N ( 5 ) -unsubstituted 1,5-dihydroflavin derivative in aque-ous solution5 . The light absorption characteristics of a few flavin-triphenylphosphine complexes are given in Table I.

The equilibrium constant of reaction ( l ) was determined spectrophotometrically for a few flavi-nium salts in acetonitrile solutions.

(C 6 H 5 ) 3 P + C104

Thus, by analysis of the extent of bleaching of the flavin solution under investigation by various tri-phenylphosphine concentration, the equilibrium con-centration of flavin, triphenylphosphine and adduct can be determined and K, the dissociation constant, calculated.

K _ [ f l a v i n ] [P(C 6 H 5 ) 3 ] [fiavin-P (C6H5) 3-adduct]

The results are summarized in Table I. Similar re-sults were obtained when the spectral data were treated a c c o r d i n g to BENESI a n d HILDEBRAND 8 .

The rate of the formation of the flavin-triphenyl-phosphine complex was found to be of pseudo first order. When, e.g. 4 x 1 0 - 5 acetonitrile solution of the flavin derivative shown in eqn. ( I ) was reacted with 4 x 1 0 - 2 M, 2 x 10~ 2 M or 4 x 1 0 - 3 M triphe-nylphosphine in acetonitrile, plots of log (A — A f in ; ii) at 380 nm vs time were linear. The rate constants were, respectively, 1.03 m i n - 1 , 0.51 m i n - 1 and 0.10 m i n - 1 at these triphenylphosphine concentra-tions. In Table I are also listed the second order rate constants of the triphenylphosphine complexes of various flavinium salts.

Wavelength (nm)

FLAVIN-PHOSPHINE ADDUCTS

Table I. Some properties of various flavin-phosphine adducts a .

1025

Complex x m a x e K c k d

(nm) (mM-1 • c m - 1 )

CH3

C H 3 Ä N \ N I OC2H5

( C 6 H 5 ) 3 - P + C I 0 4

CH3

CH3

CH 2 -CH 2 I I

I 0 ( C 6 H 5 ) 3 - P + CI04

CH3

CH3

CH3

CH3 CH3

N̂ „ 0

( C 6 H 5 ) 3 - P + CI04

rpoxt ( C 6 H 5 ) 3 - P + CI04

295 (sh)b

277 (sh) 225

328 298 (sh) 225

325 295 225

7.60 9.75

41.0

3.40 4.15

26.2

4.78 5.55

35.8

too instable to CH3 be measured

8.55 X 10-5 M

1.38 X 10-4 M

3.05 X lO"3 M

40.9 M"1 min - 1

5.70 X 10-4 ai 25.6 M-1 min-1

14.7 M_1 min - 1

8.50 M_1 min - 1

a All measurements were done with acetonitrile solutions at I phosphine complex formation rate constant.

Support for the proposed structure of the flavin-triphenylphosphine complex shown in eqn. (1) comes from elemental analysis, infrared and nuclear magnetic resonance data obtained from the crystal-line compound. Thus, the elemental analysis re-vealed that one mole triphenylphosphine is bound per mole flavin. The most characteristic absorptions of the infrared spectrum of the crystalline complex (using KBr pellet technique) are located at 3125, 2940, 1724, 1649 and 1096 cm" 1 , respectively, and are assigned to N H ( 3 ) , C O ( 4 ) , CO(2) and C104". The corresponding absorptions of the parent com-pound of the triphenylphosphine complex appear at 3200, 3035, 1750, 1710 and 1100 cm" 1 . The nuclear magnetic resonance spectrum of the com-plex in deuterated dimethylsulfoxide (internal standard was tetramethylsilane) exhibits the fol-lowing absorptions at 7.71 and 7.26 ppm (15H, tri-phenyl residue), 6.73 and 6.23 ppm (2H, CH-aro-matic), 4.68 ppm (2H, N C H 2 - 1 ) , 4.12 ppm (2H, N C H o - 1 0 ) , 2.03 ppm (3H, C H 3 - 8 ) , 1.74 ppm (3H, CH3 — 7 ) . These results are in agreement with the postulated structure.

The above described results show that flavins ex-hibit a greater reactivity towards sulfite than to-

°C. b sh = shoulder, c K = dissociation constant, & k =

wards triphenylphosphine, i .e . the neutral flavin molecule forms complexes with sulfite but not with triphenylphosphine. The protonated, respectively, the W-l-alkylated flavin species reacts about 10 times faster with sulfite than with triphenylphos-phine. The lower reactivity of flavins towards tri-phenylphosphine is partly due to the bulkiness of the phosphine molecule.

In a previous paper5 , we have shown that a linear relationship exists between the oxidation-reduction potential of the flavin derivatives and their reactivity towards sulfite. A similar relation-ship exists probably also for the triphenylphosphine-adduct reaction. This relationship reflects the elec-tron density at the flavin "active site", i .e . N ( 5 ) , C(4a) double bound, where the covalent addition (nucleophilic attack) occurs. To explain the dif-ference in reactivity of flavoproteins towards sulfite4 we have postulated that the flavoprotein oxidases possess a positively charged group in the vicinity of the prosthetic group, whereas the flavo-protein dehydrogenases possess a negatively charged group. Support for this idea comes from several papers where evidence is presented for anion binding sites of various flavoprotein oxidases such as

FLAVIN-PHOSPHINE ADDUCTS 1026

D-amino acid oxidase9 '10, L-amino acid oxidase11

and glycollate oxidase 12. The different reactivity of flavoprotein towards

sulfite could be explained postulating for the flavo-protein oxidases, the positively charged group be located close to the N ( l ) atom which is the most basic center of the prosthetic group 13 (cf. Scheme A, where X might be S or NH 2 ) . The degree of the hydrogen bridge formation would thus be one among other factors determining the reactivity of the protein towards nucleophiles, e.g. coenzyme-

substrate interaction. A possible interaction of the substrate with the N (5)-atom of the prosthetic group requires in plane interaction. Therefore, the

1 H. KAMIN, ed., Flavins and Flavoproteins, University Park Press, Baltimore 1971.

2 P . HEMMERICH, G . NAGELSCHNEIDER, a n d C . VEEGER, FEBS Letters 8, 69 [1970].

3 F. MÜLLER and V . MASSEY, i n : Methods in Enzymology, V o l . 1 8 B , p . 4 6 8 , E d s . D . B . M C C O R M I C K a n d L . D . WRIGHT, Academic Press, New York 1971.

4 V . MASSEY, F . MÜLLER, R . FELDBERG, M . SCHUMAN, P . A . SULLIVAN, L . G . H O W E L L , S . G . M A Y H E W , R . G . M A T T H E W S , a n d G . P . FOUST, J . b i o l . C h e m i s t r y 2 4 4 , 3 9 9 9 [ 1 9 6 9 ] .

5 F . MÜLLER a n d V . MASSEY, J . b i o l . C h e m i s t r y 2 4 4 , 4 0 0 7 [ 1 9 6 9 ] .

6 B . E . P . SWOBODA a n d V . MASSEY, J . b i o l . C h e m i s t r y 2 4 1 , 3409 [1966].

positively charged group at N ( l ) of the prosthetic group is not identical with the substrate binding site, but it might well be involved in the binding of inhibitors. Thus, binding of e. g. benzoate to D-amino acid oxidase to the positively charged group placed close to N ( l ) - a t o m of the prosthetic group would allow a maximum of interaction be-tween the 7i-system of the coenzyme and that of the inhibitor which manifests in the pertubation of the flavin light absorption spectrum 9 ' 1 0 . Studies con-cerning the possible interaction of flavoproteins with phosphine derivatives are now in progress.

I am indebted to Dr. MASSEY for his interest in this work. This investigation was supported by a grant from the United States Public Health Service (GM 11106), given to Dr. MASSEY.

This investigation has been performed during the tenure of a Career Development Award from the National Institutes of Health, U.S. Public Health Ser-vice ( K 4 - G M 4 2 . 5 9 9 ) .

7 W . H . W A L K E R , P . HEMMERICH, a n d V . MASSEY, E u r . J . Biochem. 13,258 [1970].

8 H . A . BENESI a n d J . A . HILDEBRAND, J . A m e r . c h e m . S o c . 71, 2703 [1949].

9 K. YAGI and T. OZAWA, Biochim. biophysica Acta [Amster-dam] 56, 420 [1962].

10 V. MASSEY and H. GANTHER, Biochemistry 4 , 1161 [ 1 9 6 5 ] . 11 A. DE KOK and C. VEEGER, Biochim. biophysica Acta [Am-

sterdam] 167, 35 [1968]. 12 M. SCHUMAN and V. MASSEY, Biochim. biophysica Acta

[Amsterdam] 227, 500, 521 [1971]. 1 3 K . H . DUDLEY, A . EHRENBERG, P . HEMMERICH, a n d F .

MÜLLER, Helv. chim. Acta 47,1354 [1964].