Formation of copper-bleomycin complexes: Evidence of a three-step process

Formation of Copper-Bleomycin Complexes: Evidence of a Three-Step Process

Jean-Paul Albertini and Arlette Garnier-Suillerot Laboratoire de Chimie Bioinorganique, U.E.R. de Mddecine et Biologie Humaine, Universit6 Paris Nord, Bobigny, France and D6partement des Recherches Physiques, Universitd Pierre et Marie Curie, Paris, France

A B S T R A C T

The formation of Cu(II)-bleomycin complexes as a function of pH has been studied using circular dichroism, absorption, electron paramagnetic resonance spectroscopy, and potentiometric titration. Our data support the following points: the formation of Cu(II)- bleomycin complexes occurs in a three-step process: a first complex (I) is formed at pH 1.2, which most probably involves the pyrimidine nitrogen, the secondary amine nitrogen, and two water molecules as the four in-plane ligands of copper. A second complex (H) is formed at pH 2.5, through the further coordination of the peptide nitrogen of histidine residue, and histidine imidazole nitrogen giving rise to the release of two protons. The fixation, in apical position, of the o~-amino nitrogen of ~-aminoalanine occurs in a last step through the release of one additional proton. A value of 2.7 has been obtained for the pK of formation of this third complex, which is the species present at physiological pH. In the Cu(II)- depbleomycin system only one complex (II') has been detected.

I N T R O D U C T I O N

Bleomycins are a family of glycopeptide antibiotics cl inically prescribed for the t reatment of selected neoplastic deseases [1]. Their therapeutic and cytotoxic activity is at tr ibuted to their ability to cleave DNA [2-4]: b leomycins induce a degradat ion of DNA in a reaction that has been shown to depend, in vitro, on the presence of ferrous ion and molecular oxygen [5-7]. Cu -b l eomyc in is inactive in

the DNA strand scission reaction, but preformed C u - b l e o m y c i n is an active ant i tumor agent [8-11]; it is now well documented that most of the b leomycin

Address reprint requests to Dr. Arlette Garnier-Suillerot, Laboratoire de Chimie Bioinorganique, U.E.R. de Medecine et Biologie Hnmaine, Universit~ Paris Nord, 74 Rue Marcel Cachin, 93012 Bobigny, France.

Journal of lnorganic Biochemistry 25, 15-24 (1985) © 1985 Elsevier Science Publishing Co., Inc. 52 Vanderbflt Ave., New York, NY 10017

15 0162-0134/85/$3.30

1 6 J.-P. Albertini and A. Garnier-Suillerot

binds to the cupric ion in blood or body fluid and is excreted as its copper complex. In order to explain the discrepancy between the in vitro and in vivo activity of the Cu(II) complex, Umezawa and co-workers [12] hypothesized that in vivo Cu(II) is reduced and then released, and the free bleomycin is then able to bind iron and activate oxygen.

Up to now, the x-ray crystallographic analysis of Cu(II)-bleomycin has not been reported; nevertheless, an x-ray structure determination of a bleomycin fragment, P-3A, shows cupric ion in square pyramidal coordination with the secondary amine nitrogen, pyrimidine ring nitrogen, deprotonated peptide nitrogen of histidine residue, and histidine imidazole nitrogen coordinate as the basal planar donor and the o~-amino nitrogen of/%aminoalanine coordinated as the axial donor [101.

Cu-bleomycins have been the subject of extensive investigation. Binding of the imidazole moiety of bleomycin to cupric ion was identified using electron spin- echo envelope spectroscopy [13]. Recent EPR study has shown that three approximately equivalent and one substantially unequivalent nitrogen donor atoms are bound to bleomycin in a square plane of a configuration which may be pyramidal square planar [14]. However, alternative structures have been proposed involving other atoms [15, 16]. Studies have already been performed showing that as a function of pH, two Cu-bleomycin complexes are formed [17].

In this paper we report potentiometric and spectroscopic data showing that in fact, as a function of pH, three distinct Cu(II)-bleomycin complexes can be identified: a complex I at very acid pH (about 1), a complex II at pH 2.5, and a complex III, which is fully formed at pH 4. Several lines of evidence suggest that the conversion of c6.aplex II to III is achieved by the binding of the c~-amino group of the/~-aminoalanine portion in apical position. This behavior compares with that of the Fe(III)-bleomycin complex which converts from a high-spin to low-spin species through binding of this o~-amino group [18, 19].

M A T E R I A L S A N D M E T H O D S

Purified bleomycin A2, which contains (3-aminopropyl) dimethylsulfonium [-NH(CH2)3S÷(CH2)2] at the terminal amine, and depyruvamide bleomycin A2 (depbleomycin A2) were kindly provided by the Laboratoire Roger Bellon. Standard Cu(II) solutions were prepared from reagent grade Cu(CIO4)2. All other reagents were of the highest quality available, and deionized distilled water was used throughout the experiments.

Unless otherwise stated the concentrations of the working solutions were 10 -3 M in bleomycin and in cupric ions. KCI 0.05 M was added. Cu(II)-bleomycin complex was formed by mixing equimolar quantities of bleomycin Cu(CIO4)2 in deionized distilled water.

Absorption spectra were recorded on a Cary 219 spectrophotometer, circular dichroism (CD) spectra on a Jobin Yvon dichrograph model Mark V, and electron paramagnetic resonance (EPR) spectra on a Varian CSE 109 spectrophotometer at - 180"C. Potentiometric measurements were obtained with a Tacussel pH meter, model Isis 20 000, at 25"C under a nitrogen atmosphere with a Metrohm EA 147 glass electrode.

Bleomycin-Copper Complexes, Three-Step Formation 1 7

RESULTS

Spectrophotometric Titration of Cu(ID-Bleomycin

Figure 1 shows the CD and absorption spectra of a Cu(II)-bleomycin solution as a function o f p H . The pH value is varying from 0.5 to 8. From pH 0.5 to 1.1 one observes the formation of a complex hereafter labeled I; the CD and absorption spectra exhibit bands at 700 and 640 nm, respectively. In the uv region the CD spectrum compares with that of free bleomycin at the same pH value. When the pH is increased from 1.1 to 2.5 the CD spectrum is completely modified: positive bands appear at 550 and 282 nm and a negative one at 650 rim. The absorption band shifts to 595 nm; it is accompanied by an isosbestic point at 665 nm. At pH 2.5 this complex (II) is fully formed. A further increase of pH from 2.5 to 4 does not give rise to spectacular modification of the absorption in the visible: the band slightly moves from 595 to 605 nm with almost the same molar absorption coefficient. Nevertheless, an isosbestic point is detectable at 625 nm. At the opposite, the change in the CD spectrum is striking: at pH 4, i .e., when this third complex (III) is fully formed, the CD spectrum in the visible region exhibits bands at 680 and 570 nm. The transition from complex II to complex III occurs with the presence of a very well-defined isodichroic point at 650 nm. As can be seen on Fig. 1, the uv CD spectrum also undergoes striking changes; an isodichroic point is apparent at 270 nm.

FIGURE 1. Absorption (lower curves) and CD (upper curves) spectra of Cu(II)-bleomycin system at different pH values ([bleomycin] = l0 -3 M, [KCI] = 0.05 M): pH 1.2 (---), 2.5 ( . . . . ), 3.3 (-O-), 5.2 ( ).

.8

.6

.4

.2

0 "~ 0 W

-2

W

10

I/ / \ ,., ..,._ , / . . - " \ ":.. ...... . - 7

,"" X "-"/

250 300

. : . Y \b~ o.Y . - 7

500 600 700 nm

.1

.0.5

O

t.O 0

-0.5

10 ~

O


The variation of the amplitude of the dichroic bands at 282, 312, and 680 nm, as a function of pH, characterized very well the stepwise formation of the three complexes. This variation has been plotted in Figure 2. As can be seen, the transition from complex I to II occurs with a pK equal to 1.7 and that from complex II to III with a pK equal to 3.

The same titration has also been performed by EPR spectroscopy and the spectra o f the three complexes , i .e . , at pH 1.1, 2 .5 , and 4, respectively, are shown in Figure 3. The spectrum of complex III compares with that previously reported [ 17,

FIGURE 2. Spectrophotometric titration of the Cu(II)-bleomycin system ([bleomycin] = 10 -3 M, [KCI] = 0.05 M). Ae as a function of pH at 680 nm (-o-), 312 nm (---), 282 nm ( . . . . . ).

-4

-6

-12

-10

.8

.6

.4

.2

0

-2 \ \

o

\ o

-8

-10

\

\ %

\ %

%. %

(.o <Z

\ t

\ o

\

| |

3 s 7

pH


i

g

el,

I,i,,I

..4o o.J I ¶ t

g=2

. . . , ~ - ~ . / . . r . . ~ j . . . . . ~, I

,,ooo, ,j/ ' : ~ / 1 ~

- ,V-~l " V '

'...-v

Magnet ic Field P

FIGURE 3. EPR of Cu(II)-bleomycin compiexes: pH 1.2 complex I ( .... ), pH 2.5 complex (ID ( pH 4 complex III ( - - - ) ([bleomycin] = 10 -3 M, [KCI] = 0.05 M).

20]. The EPR parameters for these complexes are as follows: gu, 2.25; All, 170 G for complex I, gll, 2.30; All, 178 Gfor complex II, and gll, 2.21; All, 178 G for complex III. It must be pointed out that similar results are obtained whether the titration is performed by decreasing or by increasing the pH in the range 1-7.

Potentiometric Titration of Cu(ll)-Bleomycin

We first performed a titration of bleomyc~n in the absence of cupric ion. The pH of the solution was first decreased from about 5.5 to 2 by addition of HC1 and the solution was titrated by addition of NaOH. In figure 4, the number o f protons ti released per bleomycin has been plotted as a function of pH (a has been calculated classicaly as ([H +] + [Na+],dded-[C1-]ioitiar[OH-])/[BLM] [21]. AS can be seen, two protons are released between pH 2 and 9. This occurs with a pKo of 4.7 and 7.5. In a previous work [17] values of 4.9 and 7.5 have been obtained. These have been assigned by others to the imidazole and o~-amino groups, respectively [22, 23]. When the titration is performed in the presence of cupric ions (Fig. 4, curve b) one observes that, at pH 2.5, about two protons are released with regard to free bleomycin. This strongly suggests that complex II, which is the entity present at pH 2.5, is formed through the release of two protons. When the pH is further increased to 4, i.e., when complex II is converted to III, one additional proton is released with pK of 2.9. Further increase of pH up to 8 does not reveal any other release of proton. This is consistent, as has been previously suggested [17], with the binding of a-amino and imidazole to copper.

2 0 J.-p. Albertini and A. Garnier-Suillerot

.2

+1

l e "

0

-1

f f

. . . . . .

t f

J

t t

t I

_....

/ I /

/ /

/ /

/

/

J 7

| i I I | I

2 3 4 5 6 7 8 9

pH

FIGURE 4. Potentiometric titration of bleomycin (lower curve) and Cu(II)-bleomycin system (higher curve) ([bleomycin] = 10 -3 M, [KC1] = 0.05 M). a, the number of protons released per bleomycin, as a function of pH.

Spectrophotometric Titration of Cu(II)-Depbleomycin

The depbleomycin lacks the ~-aminoalanine group which is supposed to bind to Cu(H) ion, in apical position, through the ~-amino nitrogen. The titration of the Cu(II)-depbleomycin complex was performed in the same conditions as for Cu(H)-bleomycin. The CD spectra thus obtained (Figure 5) exhibit bands at 620 and 515 rim. An absorption band is observed at 560 nm. As can be noticed the I Ael/bleomycin o f these bands increases as the pH is raised from 1.3 to 2.5. The complex thus formed will be hereafter labeled I I ' . One does not observe any strong modification when the pH is further increased up to 7. The striking feature is that, unlike the


- l

0

. 5

0

-5

i I ~%

/ / "v,..~ . . . . . . . ~ ,

2 5 0 300

/ / ',\ ,'1 ~

,7 \ '

0

sSo 88o a m

.0.5

-0.5

FIGURE 5. CD spectra of Cu(II)-depbleomycin system at different pH values ([depbleomycin] = 10 -3 M, [KCI] -- 0.05 M) : pH 1.2 ( ), 2.0 (-o-), 2.5 ( - - - ) .

Cu(II)-bleomycin system, the formation of a complex at pH higher than 2.5 is not observed. This strongly suggests that in the Cu(II)-bleomycin system the conversion of complex II to III, which occurs with the release of one proton, is due to the binding of the o~-amino nitrogen of the/~-aminoalanine to copper.

DISCUSSION

Our spectroscopic titrations unambiguously show that, as a function of pH, the complexation of Cu(II) to bleomycin occurs in a three-step process. On the other hand, the complexation of Cu(II) to depbleomycin seems to take place in only one step. Each complex exhibits well-defined CD and EPR spectra. For comparison all the spectral data are collected in Table 1.

Numerous data suggest that the structure of Cu-P3A is valid for Cu(II)- bleomycin complex at physiological pH, i.e., for complex HI. In such a case, the four in-plane ligands are the secondary amine nitrogen, pyrimidine ring nitrogen, deprotonated peptide nitrogen of histidine residue, and histidine imidazole nitrogen; one of the axial ligands is the ~-amino nitrogen of ~-aminoalanine. In the case of Fe(III)-bleomycin complex the special role played by this apical ligand is well documented; it has been shown that its coordination to Fe(III) occurs through the release of one proton with a pK of 4.6 [24], giving rise to the conversion of the spin state of the complex from a high-spin to a low-spin form [25]. In the case of Cu(II)-bleomycin complex the fixation of the ~-amino nitrogen of /3-aminoalanine, in a last step, is less obvious since this does not give rise either to spin state change or to striking modification of the absorption spectrum. Nevertheless the present studies shows that (i) one proton is released with a pK of 2.9, and (ii)


TABLE 1. Absorption and CD Data for Bleomycin, Cu(H)-Bleomycin Complexes: I, II, III, and Cu(II)-Depbleomycin Complex II'

Absorption CD

Bleomycin 310 8000 315 - 0.07

(pH 7) 287 14000 285 + 1.65

235 (sh) 19200 237 - 1.0

230 - 1.4

Complex 640 65 680 + 0.11

I 310 (sh) 8600 315 - 0 . 0 3

287 14000 245 -- 5.3

235 (sh) 20300 230 - 3.5

II 595 110 650 -- 0.30

550 + 0.30

325 (sh) 3200

310 (sh) 10600 282 + 6.3

290 16800 245 - 2.0

242 20300 230 - 3.5

[ ] 605 100 680 -- 0.5

570 + 1.4

325 (sh) 3900 312 - 4 . 5

310 (sh) 11800 270 (sh) + 4 . 5

292 17500 253 + 7.7

244 22000 230 - 2

II' 560 100 620 - 0.3

515 + 1.0

325 3000 320 - 0.4

310 (sh) 13000 280 + 4.3

292 20000 247 - 5.6

240 24000 230 - 7.5

strong modifications of the CD spectrum occur in this pH range in the case of Cu(II)-bleomycin but not in that of Cu(II)-depbleomycin. From these results we can infer that that is due to the coordination of the u-amino nitrogen of 8- aminoalanine in apical position.

In the fol lowing we will note BLM.Cu(II) .otNH2 as the complex III and BLM" Cu(II) as the complex II. The constant of association K of the c~-amino nitrogen to BLM.Cu(II) can be evaluated from the following equilibrium:

BLM • CU(II)+"ctNH2" ~ BLM • Cu(II) • otNH2 I I III

where "ctNH2" stands for the t~-amino nitrogen of the ~-aminoalanine portion of bleomycin. The concentration of complex II and III have been evaluated from CD and potentiometric titrations data. A pK, of deprotonation of the a-amino nitrogen of 7.5 (this work) has been used to calculate the concentration of " a N H , " . A value of 6 ___ 1 × 10 7 is obtained for K, where

[BLM • Cu(II) • otNH2] K =

[BLM • C u ( I I ) ] [ " c t N H 2 " l


It should be noticed that the value of this constant is only 106 in the case of the Fe(III)-bleomycin complex [24].

It is highly likely that, in complex II, the a-amino nitrogen in apical position is substituted by a water molecule. The position and the intensity of the visible absorption and CD bands are consistent with d-d transition of a cupric complex with four nitrogen ligands lying at the corners of the coordination square. The symmetry of the complex is nearly D4h. We have thus assigned the positive band at 550 nm to a transition from the 3d~z, 3d~ to the 3dx2_y2 orbitals and the negative trough at 650 nm to a transition from the 3d~ 7 to the 3dx2-y2 orbitals. This assignment is made according to the fact that, owing to the f acceptor ability of the peptide group, the 3dzx and 3dyz metal orbitals become slightly bonding [26-28]. Moreover, the observation that, in the CD spectrum the low-energy band is negative and the high-energy one positive, is consistent with the presence of a six- member chelate ring in the plane of coordination. Indeed, we have previously observed that copper complexes containing only five-member chelate ligands give rise to CD spectral pattern with a positive low-energy band and a negative high- energy one. When a six-member chelate ring is present, as it is the case in complex II, a reverse CD pattern is observed [29].

The CD spectrum of Cu(II)-depbleomycin complex (II') is, in the uv region, closely related to that of complex II. However, in the visible region, the CD band of complex II ' is 60 nm shifted to the uv with regards to that of complex II. This can be taken as an indication of a stronger in-plane ligand field in the case of complex II ' . The difference between the in-plane ligands of these two complexes arises at the level of the amine which is secondary in the case of bleomycin and primary in the case of depbleomycin. The steric hindrance around the nitrogen atom is lower in the case of primary amine than in the case of secondary amine; this can account for a stronger interaction between nitrogen and copper, inducing a stronger ligand field.

The visible absorption spectrum of complex III is closely related to that of complex II. However, owing to the apical binding of the c~-amino nitrogen, a 10- rim shift of the band to lower energy is observed. This is not unexpected as Fabrizzi et al. (1976) have shown that, in copper complexes, a strengthening of the axial interaction gives rise to a weakening of the in-plane field. Among the four ligands involved in the square of coordination the secondary amine will bind to Cu(II) without any release of proton. On the other hand the pKd of deprotonation of the N-1 nitrogen of pyrimidine, which most probably binds to Cu(II), has been reported to be lower than 1.0 [30, 31]. It is thus highly likely that complex I, which is formed at pH 1.2, involved these two groups as ligands. The four positions are thus occupied by two nitrogen atoms through a five-member chelate and most probably two water molecules. This assumption is corroborated by the position of the d-d band. Then the conversion of complex I to II will occur through the simultaneous binding of peptide and imidazole nitrogen with the release of two protons.

In summary our data support the assumption that (i) complex I is formed through binding of the secondary amine nitrogen and of the pyrimidine ring nitrogen, (ii) complex II is formed through the further binding of the peptide nitrogen of histidine residue and histidine imidazole nitrogen, and (iii) complex III is formed through the ultimate binding of the o~-amino nitrogen of/~-aminoalanine in apical position. This last point is the most strongly supported by our data.


We are indebted to Laboratorie Roger Bellon, which kindly supplied B L M A:.

R E F E R E N C E S

1. s .K . Carter, in Bleomycin Status and New Developments, S. K. Carter, S. T. Crooke, and H. Umezawa, Eds., Academic Press, New York, 1978.

2. S. M. Hecht, Ed., in Bleomycin: Chemical, Biochemical and Biological Aspects, Springer- Verlag, New York, 1979.

3. H. Umeznwa, in Anticancer Agents Based on Natural Products, J. M. Cassady and J. D. Douros, Eds., Academic Press, New York, 1980, p. 147.

4. R .M. Burger, J. Peisach, and S. B. Horwitz, Life Sci. 28, 715 (1981). 5. E. A. Sausviile, J. Peisach, and S. B. Horwitz, Biochem. Biophys. Res. Commun. 73, 814

(1976). 6. E .A. Sausville, R. W. Stein, J. Peisach, and S. B. Horwitz, Biochemistry 17, 2746 (1978). 7. T. T~dta, Y. Muraoka, T. Nakatani, A. Fujii, Y. Itaka, and H. Umezawa, J. Antibiot. 31, 1073

(1978). 8. E.A. Rao, L. A. Saryan, W. E. Antholine, and D. H. Petering, J. Meal. Chem. 23, 1310 (1980). 9. H. Suzuki, K. Nagai, E. Akutsu, N. Yarned, N. Tanaka, and H. Umezawa, J. Antibiot. (Tokyo)

23, 473 (1970). 10. Y. Itaka, N. Nakamura, T. Nakatani, Y. Muraoka, A. Fujii, T. Takita and H. Umezawa, J.

Antibiot. (Tokyo) 31, 1070 (1978). 11. J .C. Dabrowi~¢, F. T. Greenaway, W. E. Longo, M. Van Hussen, and S. T. Crooke, Biochim.

Biophys. Acta 517, 317 (1978). 12. K. Takahashi, O. Yoskioka, A. Matsuda, and H. Umezawa, J. Antiobiot. (Tokyo) 30, 861 (1977). 13. R.M. Burger, H. D. Adler, S. B. Horwitz, W. B. Mires, and J. Peisach, Biochemistry 20, 1701

(1981). 14. W.E. Antholine, J. S. Hyde, D. H. Sealy, and D. M. Petering, J. Biol. Chem. 259, 4437 (1984). 15. N.J . Oppenheimer, L. O. Rodriguez, and S. M. Hecht, Biochemistry 18, 3439 (1979). 16. N.J . Oppenheimer, L. O. Rodriguez, and S. M. Hecht, Proc. Natl. Acad. Sci. USA 76, 5616

(1979). 17. D. Solaiman, E. A. Rao, W. Antholine, and D. H. Petering, J. lnorg. Biochem. 12,201 (1980). 18. R.M. Burger, J. Peisach, W. E. Blumberg, and S. B. Horwitz, J. Biol. Chem. 254, 10906 (1979). 19. Y. Sugiura, J. Am. Chem. Soc. 102, 5208 (1980). 20. K. Ishizu, S. Murata, K. Miyoshi, Y. Sugiura, T. Takita, and H. Uzemawa, J. Antibiot. (Tokyo)

34, 994 (1981). 21. J. Bjerrum, in Metal Amine Formation in Aqueous Solution, Haase, Copenhagen, (1941). 22. D.M. Chen, B. I. Hawkins, and J. D. Glickson, Biochemistry 16, 2731 (1976). 23. H. Umezawa, Biomedecine lg, 459 (1973). 24. J .P . Albertini and A. Garnier-Suillerot, Biochemistry 23, 47 (1984). 25. J .C. Dabrowiak, F. T. Greenaway, F. S. Santillo, S. T. Crooke, and J. M. Essery, ACS Symp.

Sci 140, 237 (1980). 26. A. Gamier and L. Tosi, Biopolymers 14, 2247 (1975). 27. C .V. Phan, L. Tosi, and A. Gamier, Bioinorg. Chem. 8, 21 (1978). 28. A. Gamier and L. Tosi, Bioinorg. Chem. 8, 493 (1978). 29. A. Gamier, L. Mosoni, and L. Tosi, J. Inorg. Biochem. 13, 23 (1980). 30. T. Takita, Y. Muraoka, T. Nakatani, A. Fujii, Y. Itaka, and H. Umezawa, J. Antibiot. (Tokyo)

37, 801 (1978). 31. T. Takita, in Bleomycin Biochemical and Biological Aspects, S. M. Hecht, Ed., Springer-

Verlag, New York, 1979.

Received January 23, 1985; accepted January 29, 1985

Documents

Formation of copper-bleomycin complexes: Evidence of a three-step process