J. Biochem., 81, 495-503 (1977)
Reaction of Chlorocruorin with Heme Iron Ligands and Carbonyl Reagents1 Yutaka ORII and Nonaki WASHIO Department of Biology, Faculty of Science, Osaka University, Toyonaka, Osaka 560 Received for publication, May 29, 1976
Chlorocruorin was purified from Potamilla leptochaeta and the spectral properties of its derivatives were investigated. Fern- or ferrochlorocruorin did not exhibit a fernhemochrome or ferrohemochrome spectrum, respectively. Oxy- and carbonmonoxy-ferrochlorocruonn did show ferrohemochrome-type spectra. Ferrihemochromes were formed, however, when oxyor ferrichlorocruorin was treated with 0.02-0.05% SDS, and they were transformed to ferrohemochromes by reduction with sodium dithionite. Fernhemochrome formation was also brought about by increasing the pH of a femchlorocruonn solution to 9, or by liganding of extrinsic imidazole or cyanide to the ferric pigment. Therefore, it is apparent that at least one of the coordination positions on the heme iron in fern- and ferrochlorocruorin is vacant or occupied by a weak-field ligand. Titration studies of ferrichlorocruorin with imidazole indicated that this supposedly vacant coordination position was occupied first by the imidazole, and that the intrinsic ligand of protein origin was replaced finally at higher concentrations. The extrinsic ligands in the cyanide and imidazole complexes of ferrichlorocruorin were excluded from their coordination positions as the protein moiety assumed conformations inherent to the reduced pigment. Spectral analyses indicated that the intrinsic ligand is an imidazole moiety of a histidyl residue. When chlorocruorin was intact, carbonyl reagents such as cyanide and sodium bisulfite did not add to the formyl group of chlorocruoroheme. When the protein conformation was perturbed by SDS, addition to ferrichlorocruorin occurred appreciably. This addition was accelerated if the heme iron coordination position had been occupied by strong field ligands, and was reversed to some extent as the chlorocruorin complexes were reduced.
Chlorocruorin is a blood pigment of certain polychaete worms and undergoes reversible oxygenation as a physiological function. This pigment, like cytochrome oxidase [EC 1.9.3.1], is red in concen1
This work was supported in part by a grant from the Ministry of Education, Science and Culture of Japan. Abbreviations: SDS, sodium dodecyl sulfate; CTAB, cetyltnmethylammonium bromide. Vol. 81, No. 2, 1977
495
trated solution and green in dilute solution. This dichroic property was observed by Fox in 1926 (7); he also observed that the absorption bands of oxyand ferrochlorocruorin were very similar to those of corresponding derivatives of hemoglobin but that the ^ n d s were shifted to longer wavelengths. Further, he noted that a ferrohemochrome prepared by adding alkali to dithionite-reduced oxychlorocruorin exhibited an unstable spectrum: the initial
496
one, with peaks at 584 and 537 nm, changed to one with peaks at 569 and 533 nm with time. In the 1920's Keilin (2) observed similar phenomena with cytochrome a, and accordingly he reasoned that the prosthetic group of cytochrome a closely resembled that of chlorocruorin. Later, this spectral change was explained in terms of Schiff base formation by Lemberg (3). On the other hand, Warburg (4) indicated that the Soret band of a respiratory enzyme (=cytochrome oxidase) -CO complex was located at 433 nm, close to that of either chlorocruorin-CO at 439 nm or spirographishemoglobin-CO at 434 nm. Thus, long before the establishment of the chemical structure of dichroic heme a, the prosthetic group of cytochrome oxidase, in 1963 (5), its chemical properties had been inferred from those of chlorocruoroheme, the structure of which was determined in 1936 (6). Thus, both cytochrome oxidase and chlorocruorin contain hemes having a formyl side chain which is vulnerable to the attack of carbonyl reagents. It has been shown that a heme a-diimidazole complex forms cyanohydrin with cyanide more readily in the oxidized than the reduced state (7). Cyanohydrin as well as Schiff base formation with cytochrome oxidase, however, was suggested to occur more easily in the reduced state (8). A recent investigation further indicated possible hydrogen bond formation between the formyl group and a perturbing solvent molecule (9), and the formation of a photodissociable bond (10) between the formyl group and a certain surrounding amino acid residue when cytochrome oxidase in the reduced state was treated with alkali and brought to cryogenic temperatures. It is of interest to know whether such phenomena are unique to cytochrome oxidase or common among hemoproteins with hemes having the formyl group. In the present investigation, we have tried to clarify the nature of the heme ligand(s) of chlorocruorin as well as the reactivities of the formyl group in an attempt to extend comparatives studies.
Y. ORH and N. WASfflO Green blood taken from the base of the crown of the worms was diluted with 0.1% (w/v) NaCl solution and centrifuged for 20 min at 10,000 X g to remove insoluble materials. The supernatant was made up to 0.4 saturation with pulverized ammonium sulfate while the pH was kept at 7.4 by addition of an ammoniacal solution. After standing overnight, precipitates, if any, were removed by centrifugation for 20min at 10,000xg. The supernatant was made up to 0.9 saturation with respect to ammonium sulfate and again allowed to stand overnight. The precipitates were collected by centrifugation and dissolved in 50 HIM sodium phosphate buffer, pH 7.4. The clear solution was centrifuged for 120 min at 220,000X0, and the pellets of chlorocruorin at the bottom of the centrifuge tubes were dissolved in the same buffer. This procedure was repeated once, yielding the final preparation. In this preparation, chlorocruorin exists almost completely in the oxygenated form. In one case 2% of the total pigment was in the oxidized state and 92% in the oxygenated state. The remaining portion corresponded to the reduced (deoxygenated) form. The purified preparation was stored in an ice box for immediate use, or frozen and kept in liquid nitrogen for prolonged storage. Preparation of Ferrichlorocruorin—Ferrichloro-
cruorin was prepared by passing a solution of chlorocruorin, which had been oxidized with a small amount of solid potassium ferricyanide for 30 min, through a column of Sephadex G-25 equilibrated with 50 mM sodium phosphate buffer (pH 7.4). Other Materials—Sodium dithionite was used for either deoxygenation of oxychlorocruonn or reduction of ferrichlorocruorin, which usually took 20-30 min. Acidic or alkaline solutions of some reagents were neutralized when necessary. The concentration of chlorocruorin was determined spectrophotometrically using s m-& of 32 at 584 nm for its pyridine hemochrome (12), which was prepared in 20% (v/v) pyridine-0.1 N NaOH. For reduction of the complex, a controlled amount of sodium dithionite was used. MATERIALS AND METHODS Absorption spectra were recorded on a Cary Preparation of Chlorocruorin—Chlorocruorin model 16 spectrophotometer equipped with accesfrom Potamilla leptochaeta was prepared according sories for automatic recording. Usually recording to a modification of the method of Antonini et al. in the visible range was done on a full scale of 0.2 (11) as follows. All the operations were done at or absorbance, whereas in the Soret region the 1.0 absorbance scale was used. All measurements were below 4° unless otherwise stated. / . Biochem.
CHLOROCRUORIN DERIVATIVES
497
made at room temperature. RESULTS Absorption Spectra of Chlorocruorin—Figure shows the absorption spectra of chlorocruorin derivatives. Sodium dithionite reduced ferrichlorocruorin only slowly, and for complete reduction 20-30 min was usually required. When sodium dithionite was added to oxychlorocruorin, the pigment was deoxygenated slowly and was not converted completely into the reduced form in the same time: a small peak persisted at 600 nm. Oxychlorocruorin was prepared by bubbling pure oxygen for 1 min through a solution of dithionite-reduced chlorocruorin. Unlike ferri- and ferrochlorocruorin, the oxygenated form had a well-characterized two-banded spectrum in the visible region. CO-Chlorocruorin was prepared in the same way except that CO gas was used. Its absorption spectrum (not shown here) had distinct a- and /3-peaks at 603 and 577 nm, respectively, although the trough between them was less marked than that
for the oxygenated form. The Soret peak was at 437 nm. The spectral indices are summarized in Table I together with those given for Spirographis chloro1cruorin by other investigators for comparison. pH Dependence of Absorption Spectra—The pH dependences of the absorption spectra of ferri-, ferro-, and oxychlorocruorin were examined. When the pH was changed from 7.4 to 6.6, the absorption spectrum of ferrichlorocruorin was practically unchanged in the visible region, although the Soret peak shifted from 414 to 409 nm. Below pH 5.5, appreciable turbidity was observed. As the pH was brought up to 7.9 from neutrality, a small peak appeared at 600 nm, the Soret peak shifting to 420 nm, and at pH 8.9 a fernhemochrome-type spectrum appeared with distinct peaks at 600, 552, and 422 nm. Above that pH, the peaks in the visible region became diffuse and were finally replaced by a small bump around 610 nm at pH 10.4. The Soret peak was at 398 nm. The absorption spectrum of ferrochlorocruorin was almost unchanged between pH 5.7 and 9.3. At
125
A 100 --
^
280
75
5
A
50 -
1
605
--
20
A
i /l _ J
/i / 1
\
-— 15
1 \
i \
/
\ \
/
/
y
25 -
i
300
7
\
•'
\ '
•el
r30
/
1
400 500 WAVELENGTH (nm)
- 5
i
600
700
Fig. 1. Absorption spectra of ferri- ( ), ferro- ( ), and oxygenated ( ) chlorocruorin in 50 mM sodium phosphate buffer (pH 7.4). Fernchlorocruorin (see "MATERIALS AND METHODS") was reduced with a small amount of sodium dithionite for 20-30 min. For oxygenation, a stream of pure oxygen was passed through the reduced pigment for 1 min. Vol. 81, No. 2, 1977
498
Y. ORII and N. WASHIO
TABLE I. Comparison of the extinction coefficients of some chlorocruorin derivatives. Antonini et al. {Sp. spallanzani) Em!
Warburg; et al. (4) {Sp. spallanzani) X inm)
£mM
Chi (Fe»+)
Chi (Fet+)
Chl-O,
600 572 448 605
(14.8) (15.5) (127) (23.2)
Chi CO
438 280 602 438
570» 449
(14.3) (136.3)
(112) (73.9) (20.8)
600*
(19.6)
(130)
439
(129.8)
Scheler 03) (Sp. spallanzani) A (nm>
£ml
606 550 413
(5.6) (12.0) (97)
599 572 449 604.5 558.5 440.5
(14.0) (14.3) (143) (23.3) (13.3) (119)
598.5
(22.6)
439
(140)
Present data {P. leptochaeta) X (nm)
£mM
600 550 414 280 600 572 447 605 557 436
(11.7) (88.7) (79.1) (12.6) (12.8) (112.5) (18.6) 02.6) (92.7)
603 557 437
(18.5) (13.2) (124)
(6.78)
• Taken from Tables V and VI in Ref. 4.
pH 5.1 the solution became turbid. Between pH 9.8 and 10.1 peaks were present at 606 and 442 nm with bumps around 570 and 530 nm. Between pH 10.6 and 11.1 the only noticeable peak in the visible region was at 590 nm, the Soret peak being at 438-440 nm. At pH 11.6 a ferrohemochrome-type spectrum appeared having peaks at 584 and 436 nm. Generally the Soret peak decreased in intensity as the pH was increased from neutrality. Oxychlorocruorin was most stable to pH changes among the chlorocruorin derivatives examined. In the pH 5.5 to 11.4 region the peak positions were essentially unchanged. As the pH was lowered from neutrality, the 605-nm peak increased in intensity, whereas the 557-nm and 436-nm peaks decreased, the latter also shifting slightly towards shorter wavelength. On the other hand, as the pH was increased to 10.5, both the 605-nm and 557-nm peaks increased in intensity, whereas the 436-nm peak decreased, shifting slightly towards longer wavelength. Between pH 10.5 and 11.8 the absorption spectrum decreased in intensity throughout the wavelength region examined. Above pH 11.8 a spectrum of ferrihemochrome-type appeared. Below pH 4 the spectrum changed drastically,
probably reflecting release of the heme moiety from the protein. Profiles for these pH-dependent changes are summarized in Fig. 2. Action of Alkali on Ferrochlorocruorin—Immediately after the mixing of a solution of the dithionite-reduced pigment with an equal volume of 0.2 N NaOH a ferrohemochrome-type spectrum appeared, as shown in Fig. 3. This spectrum changed with time: the 584-nm peak decreased in intensity and was finally replaced by a new peak at 568 nm, and the 537-nm peak increased in intensity, shifting to 533 nm. All of the absorption spectra recorded within 60 min after alkahmzation had clear isosbestic points at 574, 520, and 436 nm, as shown in Fig. 3. Effects of Detergents on the Absorption Spectra —Purified Potamilla chlorocruorin is highly aggregated, like chlorocruorin obtained from Spirographis (72). Usually detergents are expected to dissociate the aggregates and consequently may alter spatial relationships between the heme moiety and its immediate environment, thus modifying the spectral properties. Therefore, the effects of some detergents on chlorocruorin at neutral pH were examined
/ . Biochem.
499
CHLOROCRUORIN DERIVATIVES
I4JO
Fig. 2. pH-Titration of ferri- (O), ferro- ( • ) , and oxygenated ( c ) chlorocruorin. The chlorocruonn derivatives (18 nmoles each) were dissolved in 3.0-ml portions of 50 rriM sodium phosphate buffer (pH 7.4). The pH of the solutions was adjusted with either 1-6 N HC1 or NaOH stepwisely. Usually the absorption spectrum was recorded 3-5 min later. The absorption change was followed at 415, 438, and 444 nm for fern-, ferro-, and oxychlorocruonn, respectively.
spectrophotometncally. When CTAB at concentrations not lower than 0.1 % (w/v) was added to a solution of oxychlorocniorin, an unstable spectrum appeared with bumps around 630 and 580 nm and a Soret peak at 407 nm, indicating partial release of heme from the protein moiety. However, when oxychlorocruorin in 0.5 % CTAB was reduced with sodium dithionite, the surviving portion showed a spectrum with a- and y9-peaks at 567 and 535 nm, respectively, and a Soret peak at 427 nm, apparently resembling the spectrum of the alkali-treated pigment, as described in the preceeding section. In the presence of CTAB at concentrations between 0.1 and 0.002%, turbidity developed accompanying loss of the spectral characteristics with time. Oxychlorocruorin solution readily became turbid on addition of a nonionic detergent, Triton X-100, at concentrations between 0.1 and 0.01% although the spectrum did not change essentially. Another nonionic detergent, Emasol 1130, did not have any effect on this solution. Unlike these detergents, an anionic detergent, SDS, affected chlorocruorin without inducing turbidity, so its effects were examined in detail. As shown in Fig. 4, the characteristic spectrum of oxychlorocruorin disappeared immediately on Vol. 81, No. 2, 1977
400 440 480 520 560 600 640 WAVELENGTH (nm) Fig. 3. Schiff base formation from fernchlorocruorin. A 30-//1 portion of oxychlorocruorin (382 //M) was dissolved in 0.75 ml of deionized water and deoxygenated with sodium dithionite for 30 min. Then, it was mixed with 0.75 ml of deaerated 0.2 N NaOH. The times at which scanning was started at 700 nm were 20 sec (characterized by the 584-nm peak), 10, 30, and 60 min (characterized by the 568-nm peak) after mixing.
addition of 0.03% SDS and was replaced by a ferrihemochrome-type spectrum similar to that observed during the alkali titration of femchlorocruorin. The absorption maxima were at 589, 545, and 422 nm. At 0.1% SDS ths 545-nm peak became smaller and shifted to 536 nm, whereas the 589-nm p:ak remained almost unchans^d, and a small bump appeared around 645 nm. This spectrum was different from that of ferrichlorocruorin, especially in the visible region, although the Soret peaks of both compounds were at 414 nm. The ferrihemochrome-type spectrum described above was also produced by the addition of SDS (0.020.03%) to ferrichlorocruorin, and when this compound was reduced with dithionite a typical ferrohemochrome-type spectrum appeared. Addition of extrinsic imidazole to an oxychlorocruorin solution treated with 0.05% SDS intensified the characteristics of the ferrihemochrome spectrum, as shown in Fig. 5. After reduction with sodium
Y. ORII and N. WASfflO
500
00 0.5 LOG(IMIDAZOLE.mM)
500 WAVELENGTH
600
700
(nm)
Fig. 4. Effects of SDS on oxychlorocruorin. Oxychlorocruorin (6.0 fiM) in 50 min sodium phosphate buffer (pH 7.4) was treated with either 0.03% (2) or 0.1% (3) SDS, and the spectrum was recorded immediately. For comparison the spectra of oxy- (1) and ferrichlorocruorin (4) are presented. OB A** 1 i OB 0.6 -
1
4 I h\ /'
04 en
3
QIO
II /Ml
-
Q2 -
QO
5SS
'• ' A / \ »5 / \ z ' " N N / sas\
/ \M V>\ \\ O05
i 400
500 WAVELENGTH
600
700
(ran)
Fig. 5. Effects of SDS plus imidazole on chlorocruorin. Oxychlorocruonn (5.4 /iM) in 50 mM sodium phosphate buffer (pH 7.4) was treated with 0.05% SDS ( ). Then imidazole (86 mM) was further added ( ). This compound was finally reduced with sodium dithionite
1.0
Fig. 6. Titration of ferrichlorocruorin with imidazole. To ferrichlorocruonn (6.22 nmoles/ml) in 50 mM sodium phosphate buffer (pH 7.4), imidazole solution (1-100 DIM) was added stepwise in 10-//1 portions. The spectral change was assumed to be complete at 17 mM imidazole. The final pH was 7.85. dithionite this compound gave a ferrohemochrome spectrum. Effects of Heme Iron Ligands—When 29 mM
imidazole was added to ferrichlorocruorin at pH 7.63 a ferrihemochrome-type spectrum similar to that in Fig. 5 appeared with absorption maxima at 595, 546, and 427 nm. The formation of this complex was analyzed by plotting log(l —a)/a against the concentration of imidazole added to the solution, where a represents the degree of complex formation estimated from the absorbance increase at 596 nm. In the lower concentration range of imidazole, the plot showed a slope of 1, whereas in the higher range the slope was 2 (Fig. 6). The shape of the absorption spectra, however, remained essentially the same during the titration. The implications of this finding will be discussed later. When the imidazole complex was reduced with sodium dithionite, a spectrum characteristic of ferrochlorocruonn appeared. The extrinsic imidazole which had been coordinated to the heme iron would have been expelled as chlorocruorin assumed the molecular conformation inherent to the reduced state. The ferrihemochrome spectrum also appeared when cyanide was added to ferrichlorocruorin, having absorption maxima at 605, 557, and 437 nm. /. Biochem.
CHLOROCRUOR1N DERIVATIVES
501
Q20
uu
1 |
\ \
/a
o
0
Q6
S 5OI2 -
B
i x
535^° rSsse
QIO
,
gO.4 in
0.08
I
1 5 CYANIDE
i
1
10
15
9 20
t/iM)
Fig. 7. Titration of ferrichlorocruorin with cyanide. To ferrichlorocruorin (11.3 nmoles/ml) in 0.2 M sodium phosphate buffer (pH 7.4), cyanide solution was added stepwise in 2-fil portions. The final pH was 7.35. The affinity of cyanide for femchlorocruorin was so strong that the latter was titrated stoichiometrically, as shown in Fig. 7. The absorbance increase at 605 nm of 11.3/JM ferrichlorocruorin reached the maximal level at 8.6 fitt cyanide, which was lower than the pigment concentration by 24%. This discrepancy may be due either to a high value for the chlorocruorin concentration, estimated based on a millimolar extinction coefficient given by Antonini et al. (12), or to aggregation of the pigment. When the cyanide complex was reduced, a typical ferrohemochrome spectrum appeared at first, having absorption maxima at 600, 549, 524, and 452 nm. This spectrum, however, changed to that of ferrochlorocruorin with time as the cyanide was expelled, and all of the spectra recorded shared clear isosbestic points at 580, 557, and 449 nm. This spectral change took longer than 60 min when the cyanide concentration was 100 fiM at neutral pH. Effect of Cyanide as a Carbonyl Reagent— When cyanide was added to SDS-treated ferrichlorocruorin, an absorption spectrum appeared having maxima at 543 and 419 nm and bumps around 600 and 670 nm. On reduction, a complicated spectrum appeared with maxima at 594, 564, 536, and 430 nm and bumps around 670 and 450 run. The ferrous sample exhibited a clear three-banded spectrum in the visible region, the maxima being at 590, 558, and 535 nm. The Soret peak was at 425 nm, flanked by a swell towards longer wavelength (Fig. 8). The 590-nm peak increased in intensity with time, whereas the other peaks decreased. As reported previously (7), these spectral characteristics apparently indicate that cyanide adds to the Voi. 81, No. 2, 1977
* i •"
Q2
1 400
/
\
M u \ V
-T Q05
/
500 WAVELENGTH
530
v
v \ •\
\
A i 600
700
(nm)
Fig. 8. Effect of cyanide on SDS-treated chlorocruorin. Oxychlorocruorin (5.4 fiM) in 50 mM sodium phosphate buffer (pH 7.4) was treated with 0.02% SDS ( ). Then cyanide (87 mM) was further added ( ) together with imidazole (20 mM). This complex was finally reduced with sodium dithionite ( ). The final pH was 7.64.
formyl group of chlorocruoroheme in both the oxidized and reduced states, although the compound is more stable in the oxidized state. The addition was accelerated when SDS-treated chlorocruorin was complexed with added imidazole. Effect of Sodium Bisulfite—Sodium bisulfite was effective in deoxygenating oxychlorocruorin but failed to reduce ferrichlorocruorin. On addition of this reagent to oxychlorocruorin treated with 0.02 % SDS its ferrihemochrome spectrum gradually changed, showing only a sharp Soret peak at 392 nm. In the visible region the bumps were at around 600 and 530 nm. When reduced with sodium dithionite, a two-banded spectrum with a- and /9-peaks at 578 and 535 nm appeared with a small bump between the peaks: the Soret peak was at 412 nm with a subsidiary one at 435 nm. On the other hand, when sodium bisulfite was added to SDS-treated ferrichlorocruorin in the presence of 20-50 mM imidazole, the ferrihemochrome spectrum changed rapidly to that of ferriproto- or ferrihematoporphyrin complex with peaks as 532 and 406 nm (Fig. 9). On reduction, three distinct peaks appeared at 582, 551, and 521 nm in the visible region and two peaks at 436 and 416 nm in the Soret. The peaks at 551, 521, and 416 nm are presumably due to an addition com-
502
Y. ORII and N. WASHIO
at about 9. The absorption maxima were at 600, 552, and 422 nm. Scheler (13) observed that the i\ >Ai transformation of an acidic form of femchloroi !i/ lt cruorin to an alkaline form proceeded in two steps Q6 i between pH 7.5-9.5 and pH 9.5-10. To the alkaline 582 i '\ 1 551 « form which appeared in the initial step he assigned a \ 1«8 Q I O spectrum with maxima at 606, 557, and 436.5 nm 'v\ (pH 9.34), differing from the above ferrihemoM chrome spectrum. Scheler ascribed the appearance \\ of this spectrum to the ionization of water coordi\\ l A Q05 \ \ \ nated to the heme iron in ferrichlorocruorin. \\ O2^ However, we could not confirm the appearance of Scheler's spectrum during pH titration, but noted that the absorption spectrum of a ferrichlorocruoi 1 0.0 700 400 500 600 rin-cyanide complex was very similar to Scheler's, WAVELENGTH (nm) the absorption maxima of the former being at 605, Fig. 9. Effect of sodium bisulfite on SDS-treated chlo- 557, and 437 nm. Oxy- or ferrichlorocruorin also rocruorin. Oxychlorocruorm (5.4 fiM) in 50 i w sodium gave a fernhemochrome spectrum on treatment with phosphate buffer (pH 7.4) was treated with 0.02% SDS SDS at neutral pH (peaks: 589, 545, and 422 nm), ( ). Then sodium bisulfite (87 ITIM) was further or when extrinsic imidazole was liganded to added ( ) together with imidazole (42 nw). This ferrichlorocruorin (peaks: 595, 546, and 427 nm). complex was finally reduced with sodium dithionite Therefore, it can be concluded that, as the imme( ). ThefinalpH was 7.17. diate environment around the heme of ferrichlorocruorin was perturbed by alkali or SDS treatment, pound, whereas the others are those of a ferrochlo- an imidazole-like ligand derived from the protein rocruorin-irrudazole complex. The formation of moiety was coordinated to the heme iron. the addition compound in the oxidized and reduced Titration studies of ferrichlorocruorin with states was less apparent when the imidazole concentration was either lower than 10 mM or higher than imidazole (Fig. 6) support the above conclusion. A 1-to-l addition of the ligand to the pigment in the 50 mM. When ferrichlorocruorin-imidazole complex initial stage corresponds to the occupation of one was not supplemented with SDS, only slight forma- vacant coordination position on the heme iron, and the next 2-to-l addition is considered to correspond tion of the addition compound was noted. to a further replacement of the intrinsic ligand by Effects of Other Carbonyl Reagents—Hydroximidazole rather than cooperative binding of ylamine, hydrazine, cysteine, and /3-mercaptoethaimidazole. The nature of this intrinsic ligand is nol did not appear to react with chlorocruorin derivatives through the formyl side chain, unless the suggested to be an imidazole moiety of a histidyl prosthetic chlorocruoroheme had been freed from residue, because its replacement with extrinsic the protein moiety. Except for hydroxylamine, imidazole did not change the spectral characteristics they acted as usual reductants, reducing ferrichloro- and the spectrum of the final product was similar to cruorin under anaerobic conditions. When the that of the femchlorocruoroheme-diimidazole reduced sample was exposed to air, oxychlorocruo- complex, with peaks at 590, 550, and 419 nm (Orii, rin was usually formed. The reducing effect of to be published). hydroxylamine was complicated by its possible Spectral changes of dithionite-reduced chloroliganding to the heme iron. cruorin after alkalinization are very similar to those observed during the Schiff base formation of cytochrome oxidase (8,14), which was carried out DISCUSSION by the reductive fixation of heme a by borohydride As the pH of a ferrichlorocruorin solution was treatment of the protein moiety in the alkali-treated increased from neutrality a ferrihemochrome-type oxidase (14). Therefore, it is reasonable to conspectrum appeared in a narrow pH region centered clude that a Schiff base is also formed between the 08
i\ ///
'II
J. Biochem.
CHLOROCRUORIN DERIVATIVES
formyl side chain of chlorocruoroheme and an •e-amino group which is brought into juxtaposition by conformational alteration in the protein moiety of chlorocruonn, although definitive proof is lacking. This may account for the phenomenon initially observed by Fox about half a century ago (/). Takemon and King (14) speculated that the Schiff base formation of cytochrome oxidase could be ascribed to the reaction of a specific amino group with the formyl group. If this is the case, both hemoproteins may possess the same spatial relationship between these two functional groups. However, if the environments of the heme moiety in these two pigments are different as regards spatial arrangements of the reacting groups, a nonspecific interaction might be responsible for the Schiff base formation. A choice between the two possibilities cannot be made until the reacting lysyl residue is identified. Addition of cyanide or sodium bisulfite to the formyl side chain of chlorocruorin occurred only under specific conditions. When chlorocruorin was intact, there was no positive indication of addition. With ferrihemochrome obtained by treatment of oxy- or ferrichlorocruorin with SDS, these carbonyl reagents reacted to give addition compounds. However, the extent of the addition was smaller than when SDS-treated chlorocruorin was coordinated with extrinsic imidazole. Upon such liganding, the peaks of the ferrihemochrome spectrum became prominent, suggesting that the addition might be promoted by a stronger ligand field or by •distorted open structure around the heme. The •existence of an optimal concentration range of imidazole for the bisulfite addition, on the other hand, suggests that at higher concentrations the •environment of the heme is distorted, becoming unfavorable for the addition. The effect of the microenvironment inherent to either the oxidized or reduced state on the addition was also clear in
"Vol. 81, No. 2, 1977
503
that the carbonyl reagents did not add to SDStreated chlorocruorin in the reduced state, though when the addition compounds in the oxidized state were reduced they remained in the reduced state, even though temporarily. Thus, in turn, addition compound formation can be regarded as a probe to survey the changes in microenvironment. The authors would like to express their thanks to Prof. M. Yoshida and the staff of the Tarnano Marine Laboratory of Okayama University, Tamano City, for providing facilities for the collection of the worms. REFERENCES 1. Fox, H.M. (1926) Proc. Roy. Soc. London B99, 199220 2. Keilin, D. (1925) Proc. Roy. Soc. London B98, 312339 3. Lemberg, R. (1964) Proc. Roy. Soc. London B159, 429-435 4. Warburg, O. & Negelein, E. (1932) Biochem. Z. 244, 9-32 5. Grassl, M., Coy, U., Seyffert, R., & Lynen, F. (1963) Biochem. Z. 338, 771-795 6. Fischer, H. & Seemann, C.V. (1936) Hoppe-Seyler's Z. Physiol. Chem. 242, 133-157 7. Yanagi, Y., Sekuzu, I., Oni, Y., & Okunuki, K. (1972) /. Biochem. 71, 47-56 8. Orii, Y. & Okunuki, K. (1964) /. Biochem. 55, 37-48 9. Yamamoto, T. & Orii, Y. (1973) /. Biochem. 74,473479 10. Oni, Y. & Iizuka, T. (1972) Biochem. Biophys. Res. Commun. 48, 884-891 11. Antonini, E., Rossi-Fanelli, A., & Caputo, A. (1962) Arch. Biochem. Biophys. 97, 336-342 12. Antonini, E., Rossi-Fanelli, A., & Caputo, A. (1962) Arch. Biochem. Biophys. 97, 343-350 13. Scheler, W. (1958) Ada Biol. Med. Germ. 1, 280-292 14. Takemori, S. & King, T.E. (1965) / . Biol. Chem. 240, 504-513
Reaction of Chlorocruorin with Heme Iron Ligands and Carbonyl Reagents1 Yutaka ORII and Nonaki WASHIO Department of Biology, Faculty of Science, Osaka University, Toyonaka, Osaka 560 Received for publication, May 29, 1976
Chlorocruorin was purified from Potamilla leptochaeta and the spectral properties of its derivatives were investigated. Fern- or ferrochlorocruorin did not exhibit a fernhemochrome or ferrohemochrome spectrum, respectively. Oxy- and carbonmonoxy-ferrochlorocruonn did show ferrohemochrome-type spectra. Ferrihemochromes were formed, however, when oxyor ferrichlorocruorin was treated with 0.02-0.05% SDS, and they were transformed to ferrohemochromes by reduction with sodium dithionite. Fernhemochrome formation was also brought about by increasing the pH of a femchlorocruonn solution to 9, or by liganding of extrinsic imidazole or cyanide to the ferric pigment. Therefore, it is apparent that at least one of the coordination positions on the heme iron in fern- and ferrochlorocruorin is vacant or occupied by a weak-field ligand. Titration studies of ferrichlorocruorin with imidazole indicated that this supposedly vacant coordination position was occupied first by the imidazole, and that the intrinsic ligand of protein origin was replaced finally at higher concentrations. The extrinsic ligands in the cyanide and imidazole complexes of ferrichlorocruorin were excluded from their coordination positions as the protein moiety assumed conformations inherent to the reduced pigment. Spectral analyses indicated that the intrinsic ligand is an imidazole moiety of a histidyl residue. When chlorocruorin was intact, carbonyl reagents such as cyanide and sodium bisulfite did not add to the formyl group of chlorocruoroheme. When the protein conformation was perturbed by SDS, addition to ferrichlorocruorin occurred appreciably. This addition was accelerated if the heme iron coordination position had been occupied by strong field ligands, and was reversed to some extent as the chlorocruorin complexes were reduced.
Chlorocruorin is a blood pigment of certain polychaete worms and undergoes reversible oxygenation as a physiological function. This pigment, like cytochrome oxidase [EC 1.9.3.1], is red in concen1
This work was supported in part by a grant from the Ministry of Education, Science and Culture of Japan. Abbreviations: SDS, sodium dodecyl sulfate; CTAB, cetyltnmethylammonium bromide. Vol. 81, No. 2, 1977
495
trated solution and green in dilute solution. This dichroic property was observed by Fox in 1926 (7); he also observed that the absorption bands of oxyand ferrochlorocruorin were very similar to those of corresponding derivatives of hemoglobin but that the ^ n d s were shifted to longer wavelengths. Further, he noted that a ferrohemochrome prepared by adding alkali to dithionite-reduced oxychlorocruorin exhibited an unstable spectrum: the initial
496
one, with peaks at 584 and 537 nm, changed to one with peaks at 569 and 533 nm with time. In the 1920's Keilin (2) observed similar phenomena with cytochrome a, and accordingly he reasoned that the prosthetic group of cytochrome a closely resembled that of chlorocruorin. Later, this spectral change was explained in terms of Schiff base formation by Lemberg (3). On the other hand, Warburg (4) indicated that the Soret band of a respiratory enzyme (=cytochrome oxidase) -CO complex was located at 433 nm, close to that of either chlorocruorin-CO at 439 nm or spirographishemoglobin-CO at 434 nm. Thus, long before the establishment of the chemical structure of dichroic heme a, the prosthetic group of cytochrome oxidase, in 1963 (5), its chemical properties had been inferred from those of chlorocruoroheme, the structure of which was determined in 1936 (6). Thus, both cytochrome oxidase and chlorocruorin contain hemes having a formyl side chain which is vulnerable to the attack of carbonyl reagents. It has been shown that a heme a-diimidazole complex forms cyanohydrin with cyanide more readily in the oxidized than the reduced state (7). Cyanohydrin as well as Schiff base formation with cytochrome oxidase, however, was suggested to occur more easily in the reduced state (8). A recent investigation further indicated possible hydrogen bond formation between the formyl group and a perturbing solvent molecule (9), and the formation of a photodissociable bond (10) between the formyl group and a certain surrounding amino acid residue when cytochrome oxidase in the reduced state was treated with alkali and brought to cryogenic temperatures. It is of interest to know whether such phenomena are unique to cytochrome oxidase or common among hemoproteins with hemes having the formyl group. In the present investigation, we have tried to clarify the nature of the heme ligand(s) of chlorocruorin as well as the reactivities of the formyl group in an attempt to extend comparatives studies.
Y. ORH and N. WASfflO Green blood taken from the base of the crown of the worms was diluted with 0.1% (w/v) NaCl solution and centrifuged for 20 min at 10,000 X g to remove insoluble materials. The supernatant was made up to 0.4 saturation with pulverized ammonium sulfate while the pH was kept at 7.4 by addition of an ammoniacal solution. After standing overnight, precipitates, if any, were removed by centrifugation for 20min at 10,000xg. The supernatant was made up to 0.9 saturation with respect to ammonium sulfate and again allowed to stand overnight. The precipitates were collected by centrifugation and dissolved in 50 HIM sodium phosphate buffer, pH 7.4. The clear solution was centrifuged for 120 min at 220,000X0, and the pellets of chlorocruorin at the bottom of the centrifuge tubes were dissolved in the same buffer. This procedure was repeated once, yielding the final preparation. In this preparation, chlorocruorin exists almost completely in the oxygenated form. In one case 2% of the total pigment was in the oxidized state and 92% in the oxygenated state. The remaining portion corresponded to the reduced (deoxygenated) form. The purified preparation was stored in an ice box for immediate use, or frozen and kept in liquid nitrogen for prolonged storage. Preparation of Ferrichlorocruorin—Ferrichloro-
cruorin was prepared by passing a solution of chlorocruorin, which had been oxidized with a small amount of solid potassium ferricyanide for 30 min, through a column of Sephadex G-25 equilibrated with 50 mM sodium phosphate buffer (pH 7.4). Other Materials—Sodium dithionite was used for either deoxygenation of oxychlorocruonn or reduction of ferrichlorocruorin, which usually took 20-30 min. Acidic or alkaline solutions of some reagents were neutralized when necessary. The concentration of chlorocruorin was determined spectrophotometrically using s m-& of 32 at 584 nm for its pyridine hemochrome (12), which was prepared in 20% (v/v) pyridine-0.1 N NaOH. For reduction of the complex, a controlled amount of sodium dithionite was used. MATERIALS AND METHODS Absorption spectra were recorded on a Cary Preparation of Chlorocruorin—Chlorocruorin model 16 spectrophotometer equipped with accesfrom Potamilla leptochaeta was prepared according sories for automatic recording. Usually recording to a modification of the method of Antonini et al. in the visible range was done on a full scale of 0.2 (11) as follows. All the operations were done at or absorbance, whereas in the Soret region the 1.0 absorbance scale was used. All measurements were below 4° unless otherwise stated. / . Biochem.
CHLOROCRUORIN DERIVATIVES
497
made at room temperature. RESULTS Absorption Spectra of Chlorocruorin—Figure shows the absorption spectra of chlorocruorin derivatives. Sodium dithionite reduced ferrichlorocruorin only slowly, and for complete reduction 20-30 min was usually required. When sodium dithionite was added to oxychlorocruorin, the pigment was deoxygenated slowly and was not converted completely into the reduced form in the same time: a small peak persisted at 600 nm. Oxychlorocruorin was prepared by bubbling pure oxygen for 1 min through a solution of dithionite-reduced chlorocruorin. Unlike ferri- and ferrochlorocruorin, the oxygenated form had a well-characterized two-banded spectrum in the visible region. CO-Chlorocruorin was prepared in the same way except that CO gas was used. Its absorption spectrum (not shown here) had distinct a- and /3-peaks at 603 and 577 nm, respectively, although the trough between them was less marked than that
for the oxygenated form. The Soret peak was at 437 nm. The spectral indices are summarized in Table I together with those given for Spirographis chloro1cruorin by other investigators for comparison. pH Dependence of Absorption Spectra—The pH dependences of the absorption spectra of ferri-, ferro-, and oxychlorocruorin were examined. When the pH was changed from 7.4 to 6.6, the absorption spectrum of ferrichlorocruorin was practically unchanged in the visible region, although the Soret peak shifted from 414 to 409 nm. Below pH 5.5, appreciable turbidity was observed. As the pH was brought up to 7.9 from neutrality, a small peak appeared at 600 nm, the Soret peak shifting to 420 nm, and at pH 8.9 a fernhemochrome-type spectrum appeared with distinct peaks at 600, 552, and 422 nm. Above that pH, the peaks in the visible region became diffuse and were finally replaced by a small bump around 610 nm at pH 10.4. The Soret peak was at 398 nm. The absorption spectrum of ferrochlorocruorin was almost unchanged between pH 5.7 and 9.3. At
125
A 100 --
^
280
75
5
A
50 -
1
605
--
20
A
i /l _ J
/i / 1
\
-— 15
1 \
i \
/
\ \
/
/
y
25 -
i
300
7
\
•'
\ '
•el
r30
/
1
400 500 WAVELENGTH (nm)
- 5
i
600
700
Fig. 1. Absorption spectra of ferri- ( ), ferro- ( ), and oxygenated ( ) chlorocruorin in 50 mM sodium phosphate buffer (pH 7.4). Fernchlorocruorin (see "MATERIALS AND METHODS") was reduced with a small amount of sodium dithionite for 20-30 min. For oxygenation, a stream of pure oxygen was passed through the reduced pigment for 1 min. Vol. 81, No. 2, 1977
498
Y. ORII and N. WASHIO
TABLE I. Comparison of the extinction coefficients of some chlorocruorin derivatives. Antonini et al. {Sp. spallanzani) Em!
Warburg; et al. (4) {Sp. spallanzani) X inm)
£mM
Chi (Fe»+)
Chi (Fet+)
Chl-O,
600 572 448 605
(14.8) (15.5) (127) (23.2)
Chi CO
438 280 602 438
570» 449
(14.3) (136.3)
(112) (73.9) (20.8)
600*
(19.6)
(130)
439
(129.8)
Scheler 03) (Sp. spallanzani) A (nm>
£ml
606 550 413
(5.6) (12.0) (97)
599 572 449 604.5 558.5 440.5
(14.0) (14.3) (143) (23.3) (13.3) (119)
598.5
(22.6)
439
(140)
Present data {P. leptochaeta) X (nm)
£mM
600 550 414 280 600 572 447 605 557 436
(11.7) (88.7) (79.1) (12.6) (12.8) (112.5) (18.6) 02.6) (92.7)
603 557 437
(18.5) (13.2) (124)
(6.78)
• Taken from Tables V and VI in Ref. 4.
pH 5.1 the solution became turbid. Between pH 9.8 and 10.1 peaks were present at 606 and 442 nm with bumps around 570 and 530 nm. Between pH 10.6 and 11.1 the only noticeable peak in the visible region was at 590 nm, the Soret peak being at 438-440 nm. At pH 11.6 a ferrohemochrome-type spectrum appeared having peaks at 584 and 436 nm. Generally the Soret peak decreased in intensity as the pH was increased from neutrality. Oxychlorocruorin was most stable to pH changes among the chlorocruorin derivatives examined. In the pH 5.5 to 11.4 region the peak positions were essentially unchanged. As the pH was lowered from neutrality, the 605-nm peak increased in intensity, whereas the 557-nm and 436-nm peaks decreased, the latter also shifting slightly towards shorter wavelength. On the other hand, as the pH was increased to 10.5, both the 605-nm and 557-nm peaks increased in intensity, whereas the 436-nm peak decreased, shifting slightly towards longer wavelength. Between pH 10.5 and 11.8 the absorption spectrum decreased in intensity throughout the wavelength region examined. Above pH 11.8 a spectrum of ferrihemochrome-type appeared. Below pH 4 the spectrum changed drastically,
probably reflecting release of the heme moiety from the protein. Profiles for these pH-dependent changes are summarized in Fig. 2. Action of Alkali on Ferrochlorocruorin—Immediately after the mixing of a solution of the dithionite-reduced pigment with an equal volume of 0.2 N NaOH a ferrohemochrome-type spectrum appeared, as shown in Fig. 3. This spectrum changed with time: the 584-nm peak decreased in intensity and was finally replaced by a new peak at 568 nm, and the 537-nm peak increased in intensity, shifting to 533 nm. All of the absorption spectra recorded within 60 min after alkahmzation had clear isosbestic points at 574, 520, and 436 nm, as shown in Fig. 3. Effects of Detergents on the Absorption Spectra —Purified Potamilla chlorocruorin is highly aggregated, like chlorocruorin obtained from Spirographis (72). Usually detergents are expected to dissociate the aggregates and consequently may alter spatial relationships between the heme moiety and its immediate environment, thus modifying the spectral properties. Therefore, the effects of some detergents on chlorocruorin at neutral pH were examined
/ . Biochem.
499
CHLOROCRUORIN DERIVATIVES
I4JO
Fig. 2. pH-Titration of ferri- (O), ferro- ( • ) , and oxygenated ( c ) chlorocruorin. The chlorocruonn derivatives (18 nmoles each) were dissolved in 3.0-ml portions of 50 rriM sodium phosphate buffer (pH 7.4). The pH of the solutions was adjusted with either 1-6 N HC1 or NaOH stepwisely. Usually the absorption spectrum was recorded 3-5 min later. The absorption change was followed at 415, 438, and 444 nm for fern-, ferro-, and oxychlorocruonn, respectively.
spectrophotometncally. When CTAB at concentrations not lower than 0.1 % (w/v) was added to a solution of oxychlorocniorin, an unstable spectrum appeared with bumps around 630 and 580 nm and a Soret peak at 407 nm, indicating partial release of heme from the protein moiety. However, when oxychlorocruorin in 0.5 % CTAB was reduced with sodium dithionite, the surviving portion showed a spectrum with a- and y9-peaks at 567 and 535 nm, respectively, and a Soret peak at 427 nm, apparently resembling the spectrum of the alkali-treated pigment, as described in the preceeding section. In the presence of CTAB at concentrations between 0.1 and 0.002%, turbidity developed accompanying loss of the spectral characteristics with time. Oxychlorocruorin solution readily became turbid on addition of a nonionic detergent, Triton X-100, at concentrations between 0.1 and 0.01% although the spectrum did not change essentially. Another nonionic detergent, Emasol 1130, did not have any effect on this solution. Unlike these detergents, an anionic detergent, SDS, affected chlorocruorin without inducing turbidity, so its effects were examined in detail. As shown in Fig. 4, the characteristic spectrum of oxychlorocruorin disappeared immediately on Vol. 81, No. 2, 1977
400 440 480 520 560 600 640 WAVELENGTH (nm) Fig. 3. Schiff base formation from fernchlorocruorin. A 30-//1 portion of oxychlorocruorin (382 //M) was dissolved in 0.75 ml of deionized water and deoxygenated with sodium dithionite for 30 min. Then, it was mixed with 0.75 ml of deaerated 0.2 N NaOH. The times at which scanning was started at 700 nm were 20 sec (characterized by the 584-nm peak), 10, 30, and 60 min (characterized by the 568-nm peak) after mixing.
addition of 0.03% SDS and was replaced by a ferrihemochrome-type spectrum similar to that observed during the alkali titration of femchlorocruorin. The absorption maxima were at 589, 545, and 422 nm. At 0.1% SDS ths 545-nm peak became smaller and shifted to 536 nm, whereas the 589-nm p:ak remained almost unchans^d, and a small bump appeared around 645 nm. This spectrum was different from that of ferrichlorocruorin, especially in the visible region, although the Soret peaks of both compounds were at 414 nm. The ferrihemochrome-type spectrum described above was also produced by the addition of SDS (0.020.03%) to ferrichlorocruorin, and when this compound was reduced with dithionite a typical ferrohemochrome-type spectrum appeared. Addition of extrinsic imidazole to an oxychlorocruorin solution treated with 0.05% SDS intensified the characteristics of the ferrihemochrome spectrum, as shown in Fig. 5. After reduction with sodium
Y. ORII and N. WASfflO
500
00 0.5 LOG(IMIDAZOLE.mM)
500 WAVELENGTH
600
700
(nm)
Fig. 4. Effects of SDS on oxychlorocruorin. Oxychlorocruorin (6.0 fiM) in 50 min sodium phosphate buffer (pH 7.4) was treated with either 0.03% (2) or 0.1% (3) SDS, and the spectrum was recorded immediately. For comparison the spectra of oxy- (1) and ferrichlorocruorin (4) are presented. OB A** 1 i OB 0.6 -
1
4 I h\ /'
04 en
3
QIO
II /Ml
-
Q2 -
QO
5SS
'• ' A / \ »5 / \ z ' " N N / sas\
/ \M V>\ \\ O05
i 400
500 WAVELENGTH
600
700
(ran)
Fig. 5. Effects of SDS plus imidazole on chlorocruorin. Oxychlorocruonn (5.4 /iM) in 50 mM sodium phosphate buffer (pH 7.4) was treated with 0.05% SDS ( ). Then imidazole (86 mM) was further added ( ). This compound was finally reduced with sodium dithionite
1.0
Fig. 6. Titration of ferrichlorocruorin with imidazole. To ferrichlorocruonn (6.22 nmoles/ml) in 50 mM sodium phosphate buffer (pH 7.4), imidazole solution (1-100 DIM) was added stepwise in 10-//1 portions. The spectral change was assumed to be complete at 17 mM imidazole. The final pH was 7.85. dithionite this compound gave a ferrohemochrome spectrum. Effects of Heme Iron Ligands—When 29 mM
imidazole was added to ferrichlorocruorin at pH 7.63 a ferrihemochrome-type spectrum similar to that in Fig. 5 appeared with absorption maxima at 595, 546, and 427 nm. The formation of this complex was analyzed by plotting log(l —a)/a against the concentration of imidazole added to the solution, where a represents the degree of complex formation estimated from the absorbance increase at 596 nm. In the lower concentration range of imidazole, the plot showed a slope of 1, whereas in the higher range the slope was 2 (Fig. 6). The shape of the absorption spectra, however, remained essentially the same during the titration. The implications of this finding will be discussed later. When the imidazole complex was reduced with sodium dithionite, a spectrum characteristic of ferrochlorocruonn appeared. The extrinsic imidazole which had been coordinated to the heme iron would have been expelled as chlorocruorin assumed the molecular conformation inherent to the reduced state. The ferrihemochrome spectrum also appeared when cyanide was added to ferrichlorocruorin, having absorption maxima at 605, 557, and 437 nm. /. Biochem.
CHLOROCRUOR1N DERIVATIVES
501
Q20
uu
1 |
\ \
/a
o
0
Q6
S 5OI2 -
B
i x
535^° rSsse
QIO
,
gO.4 in
0.08
I
1 5 CYANIDE
i
1
10
15
9 20
t/iM)
Fig. 7. Titration of ferrichlorocruorin with cyanide. To ferrichlorocruorin (11.3 nmoles/ml) in 0.2 M sodium phosphate buffer (pH 7.4), cyanide solution was added stepwise in 2-fil portions. The final pH was 7.35. The affinity of cyanide for femchlorocruorin was so strong that the latter was titrated stoichiometrically, as shown in Fig. 7. The absorbance increase at 605 nm of 11.3/JM ferrichlorocruorin reached the maximal level at 8.6 fitt cyanide, which was lower than the pigment concentration by 24%. This discrepancy may be due either to a high value for the chlorocruorin concentration, estimated based on a millimolar extinction coefficient given by Antonini et al. (12), or to aggregation of the pigment. When the cyanide complex was reduced, a typical ferrohemochrome spectrum appeared at first, having absorption maxima at 600, 549, 524, and 452 nm. This spectrum, however, changed to that of ferrochlorocruorin with time as the cyanide was expelled, and all of the spectra recorded shared clear isosbestic points at 580, 557, and 449 nm. This spectral change took longer than 60 min when the cyanide concentration was 100 fiM at neutral pH. Effect of Cyanide as a Carbonyl Reagent— When cyanide was added to SDS-treated ferrichlorocruorin, an absorption spectrum appeared having maxima at 543 and 419 nm and bumps around 600 and 670 nm. On reduction, a complicated spectrum appeared with maxima at 594, 564, 536, and 430 nm and bumps around 670 and 450 run. The ferrous sample exhibited a clear three-banded spectrum in the visible region, the maxima being at 590, 558, and 535 nm. The Soret peak was at 425 nm, flanked by a swell towards longer wavelength (Fig. 8). The 590-nm peak increased in intensity with time, whereas the other peaks decreased. As reported previously (7), these spectral characteristics apparently indicate that cyanide adds to the Voi. 81, No. 2, 1977
* i •"
Q2
1 400
/
\
M u \ V
-T Q05
/
500 WAVELENGTH
530
v
v \ •\
\
A i 600
700
(nm)
Fig. 8. Effect of cyanide on SDS-treated chlorocruorin. Oxychlorocruorin (5.4 fiM) in 50 mM sodium phosphate buffer (pH 7.4) was treated with 0.02% SDS ( ). Then cyanide (87 mM) was further added ( ) together with imidazole (20 mM). This complex was finally reduced with sodium dithionite ( ). The final pH was 7.64.
formyl group of chlorocruoroheme in both the oxidized and reduced states, although the compound is more stable in the oxidized state. The addition was accelerated when SDS-treated chlorocruorin was complexed with added imidazole. Effect of Sodium Bisulfite—Sodium bisulfite was effective in deoxygenating oxychlorocruorin but failed to reduce ferrichlorocruorin. On addition of this reagent to oxychlorocruorin treated with 0.02 % SDS its ferrihemochrome spectrum gradually changed, showing only a sharp Soret peak at 392 nm. In the visible region the bumps were at around 600 and 530 nm. When reduced with sodium dithionite, a two-banded spectrum with a- and /9-peaks at 578 and 535 nm appeared with a small bump between the peaks: the Soret peak was at 412 nm with a subsidiary one at 435 nm. On the other hand, when sodium bisulfite was added to SDS-treated ferrichlorocruorin in the presence of 20-50 mM imidazole, the ferrihemochrome spectrum changed rapidly to that of ferriproto- or ferrihematoporphyrin complex with peaks as 532 and 406 nm (Fig. 9). On reduction, three distinct peaks appeared at 582, 551, and 521 nm in the visible region and two peaks at 436 and 416 nm in the Soret. The peaks at 551, 521, and 416 nm are presumably due to an addition com-
502
Y. ORII and N. WASHIO
at about 9. The absorption maxima were at 600, 552, and 422 nm. Scheler (13) observed that the i\ >Ai transformation of an acidic form of femchloroi !i/ lt cruorin to an alkaline form proceeded in two steps Q6 i between pH 7.5-9.5 and pH 9.5-10. To the alkaline 582 i '\ 1 551 « form which appeared in the initial step he assigned a \ 1«8 Q I O spectrum with maxima at 606, 557, and 436.5 nm 'v\ (pH 9.34), differing from the above ferrihemoM chrome spectrum. Scheler ascribed the appearance \\ of this spectrum to the ionization of water coordi\\ l A Q05 \ \ \ nated to the heme iron in ferrichlorocruorin. \\ O2^ However, we could not confirm the appearance of Scheler's spectrum during pH titration, but noted that the absorption spectrum of a ferrichlorocruoi 1 0.0 700 400 500 600 rin-cyanide complex was very similar to Scheler's, WAVELENGTH (nm) the absorption maxima of the former being at 605, Fig. 9. Effect of sodium bisulfite on SDS-treated chlo- 557, and 437 nm. Oxy- or ferrichlorocruorin also rocruorin. Oxychlorocruorm (5.4 fiM) in 50 i w sodium gave a fernhemochrome spectrum on treatment with phosphate buffer (pH 7.4) was treated with 0.02% SDS SDS at neutral pH (peaks: 589, 545, and 422 nm), ( ). Then sodium bisulfite (87 ITIM) was further or when extrinsic imidazole was liganded to added ( ) together with imidazole (42 nw). This ferrichlorocruorin (peaks: 595, 546, and 427 nm). complex was finally reduced with sodium dithionite Therefore, it can be concluded that, as the imme( ). ThefinalpH was 7.17. diate environment around the heme of ferrichlorocruorin was perturbed by alkali or SDS treatment, pound, whereas the others are those of a ferrochlo- an imidazole-like ligand derived from the protein rocruorin-irrudazole complex. The formation of moiety was coordinated to the heme iron. the addition compound in the oxidized and reduced Titration studies of ferrichlorocruorin with states was less apparent when the imidazole concentration was either lower than 10 mM or higher than imidazole (Fig. 6) support the above conclusion. A 1-to-l addition of the ligand to the pigment in the 50 mM. When ferrichlorocruorin-imidazole complex initial stage corresponds to the occupation of one was not supplemented with SDS, only slight forma- vacant coordination position on the heme iron, and the next 2-to-l addition is considered to correspond tion of the addition compound was noted. to a further replacement of the intrinsic ligand by Effects of Other Carbonyl Reagents—Hydroximidazole rather than cooperative binding of ylamine, hydrazine, cysteine, and /3-mercaptoethaimidazole. The nature of this intrinsic ligand is nol did not appear to react with chlorocruorin derivatives through the formyl side chain, unless the suggested to be an imidazole moiety of a histidyl prosthetic chlorocruoroheme had been freed from residue, because its replacement with extrinsic the protein moiety. Except for hydroxylamine, imidazole did not change the spectral characteristics they acted as usual reductants, reducing ferrichloro- and the spectrum of the final product was similar to cruorin under anaerobic conditions. When the that of the femchlorocruoroheme-diimidazole reduced sample was exposed to air, oxychlorocruo- complex, with peaks at 590, 550, and 419 nm (Orii, rin was usually formed. The reducing effect of to be published). hydroxylamine was complicated by its possible Spectral changes of dithionite-reduced chloroliganding to the heme iron. cruorin after alkalinization are very similar to those observed during the Schiff base formation of cytochrome oxidase (8,14), which was carried out DISCUSSION by the reductive fixation of heme a by borohydride As the pH of a ferrichlorocruorin solution was treatment of the protein moiety in the alkali-treated increased from neutrality a ferrihemochrome-type oxidase (14). Therefore, it is reasonable to conspectrum appeared in a narrow pH region centered clude that a Schiff base is also formed between the 08
i\ ///
'II
J. Biochem.
CHLOROCRUORIN DERIVATIVES
formyl side chain of chlorocruoroheme and an •e-amino group which is brought into juxtaposition by conformational alteration in the protein moiety of chlorocruonn, although definitive proof is lacking. This may account for the phenomenon initially observed by Fox about half a century ago (/). Takemon and King (14) speculated that the Schiff base formation of cytochrome oxidase could be ascribed to the reaction of a specific amino group with the formyl group. If this is the case, both hemoproteins may possess the same spatial relationship between these two functional groups. However, if the environments of the heme moiety in these two pigments are different as regards spatial arrangements of the reacting groups, a nonspecific interaction might be responsible for the Schiff base formation. A choice between the two possibilities cannot be made until the reacting lysyl residue is identified. Addition of cyanide or sodium bisulfite to the formyl side chain of chlorocruorin occurred only under specific conditions. When chlorocruorin was intact, there was no positive indication of addition. With ferrihemochrome obtained by treatment of oxy- or ferrichlorocruorin with SDS, these carbonyl reagents reacted to give addition compounds. However, the extent of the addition was smaller than when SDS-treated chlorocruorin was coordinated with extrinsic imidazole. Upon such liganding, the peaks of the ferrihemochrome spectrum became prominent, suggesting that the addition might be promoted by a stronger ligand field or by •distorted open structure around the heme. The •existence of an optimal concentration range of imidazole for the bisulfite addition, on the other hand, suggests that at higher concentrations the •environment of the heme is distorted, becoming unfavorable for the addition. The effect of the microenvironment inherent to either the oxidized or reduced state on the addition was also clear in
"Vol. 81, No. 2, 1977
503
that the carbonyl reagents did not add to SDStreated chlorocruorin in the reduced state, though when the addition compounds in the oxidized state were reduced they remained in the reduced state, even though temporarily. Thus, in turn, addition compound formation can be regarded as a probe to survey the changes in microenvironment. The authors would like to express their thanks to Prof. M. Yoshida and the staff of the Tarnano Marine Laboratory of Okayama University, Tamano City, for providing facilities for the collection of the worms. REFERENCES 1. Fox, H.M. (1926) Proc. Roy. Soc. London B99, 199220 2. Keilin, D. (1925) Proc. Roy. Soc. London B98, 312339 3. Lemberg, R. (1964) Proc. Roy. Soc. London B159, 429-435 4. Warburg, O. & Negelein, E. (1932) Biochem. Z. 244, 9-32 5. Grassl, M., Coy, U., Seyffert, R., & Lynen, F. (1963) Biochem. Z. 338, 771-795 6. Fischer, H. & Seemann, C.V. (1936) Hoppe-Seyler's Z. Physiol. Chem. 242, 133-157 7. Yanagi, Y., Sekuzu, I., Oni, Y., & Okunuki, K. (1972) /. Biochem. 71, 47-56 8. Orii, Y. & Okunuki, K. (1964) /. Biochem. 55, 37-48 9. Yamamoto, T. & Orii, Y. (1973) /. Biochem. 74,473479 10. Oni, Y. & Iizuka, T. (1972) Biochem. Biophys. Res. Commun. 48, 884-891 11. Antonini, E., Rossi-Fanelli, A., & Caputo, A. (1962) Arch. Biochem. Biophys. 97, 336-342 12. Antonini, E., Rossi-Fanelli, A., & Caputo, A. (1962) Arch. Biochem. Biophys. 97, 343-350 13. Scheler, W. (1958) Ada Biol. Med. Germ. 1, 280-292 14. Takemori, S. & King, T.E. (1965) / . Biol. Chem. 240, 504-513
