ARCHIVES
Vol.
OF BIOCHEMISTRY
284, No. 1, January,
AND
BIOPHYSICS
pp. 112-115,
1991
Intrinsic Fluorescence of the Bacterial CopperContaining Protein Amicyanin’ Nicola Rosato, Giampiero Mei, Isabella Savini, Arjen Lommen,* and Gerard W. Canters*
Franc0
Del Bolgia,
Dipartimento di Medicina Sperimentale & Scienze Biochimiche, Universitci *Gorlaeus Laboratories, Department of Chemistry, State University Leiden,
Received
July
Alessandro
di Roma “Tor Leiden.
Finazzi-Agrb,2
Vergata;”
and
9, 1990
The fluorescence properties of the single tryptophanyl residue present in amicyanin from Thiobacillus versutus are very similar to those of azurin from Pseudomonas aeruginosa and other mononuclear blue copper proteins. The emission maximum is well structured and centered at 3 18 nm. The quantum yield is strongly affected by the presence of copper, the removal of which is accompanied by a more than sixfold increase in fluorescence, without change in shape. The fluorescence decay of holo-amicyanin is heterogeneous with a longer component of 5.7 ns and a shorter one of 0.7 ns accounting for 90% of the total emitting molecules. Copper-free amicyanin shows instead a single exponential decay (3.3 ns) of intrinsic fluorescence. This lifetime decreases as the temperature increases as does the longer lifetime component of holoamicyanin. 0 1991 Academic Press. Inc.
Amicyanin is a member of the family of bacterial and plant mononuclear blue copper proteins (1). Its intense blue color, EPR spectrum, redox potential, copper content, molecular weight, and amino acid composition recall those of azurin, plastocyanin, plantacyanin, rusticyanin, and so on (2). The presence of a single tryptophan in amicyanin analogous to Pseudomonas aeruginosa and Pseudomonas fluorescens azurins (3) suggested to us a study of its static and dynamic fluorescence. The fluorescence of azurins often shows very peculiar spectroscopic properties, becausethe single tryptophan shows a very complicated fluorescence decay. The origin of this heterogeneous decay is still a matter of controversy (4, 6). 1 This work has been partly supported by Italian Minister0 della Pubblica Istruzione “Progetti Nazionali 40%.” ‘To whom correspondence should be addressed at Dipartimento di Medicina Sperimentale e Scienze Biochimiche, Universith di Roma “Tor Vergata,” Via 0. Raimondo, 00173 Roma, Italy; 6-7230291; Fax: 6-6130793.
In the present paper we report that amicyanin also has static and dynamic fluorescence properties very different from those of simple tryptophan model compounds but similar to those of the bacterial protein azurin. The results presented here rule out some hypotheses on the complex decay in single tryptophan-containing proteins. MATERIALS
AND
METHODS
Purification of amicyanin. The bacterial paste of Thiobacillus uersutus ATCC 25364T (l), which was used for the preparation of amicyanin, was taken from a supply grown previously (7). The isolation procedure for amicyanin was modified as follows: 200 g of wet bacterial paste was resuspended in 300 ml of IO mM phosphate buffer (pH 7.0). The cell suspension was incubated for 1 h at 15’C with 200 mg lysozyme, 180 mg EDTA, and 0.5 ml of a solution of 100 mM PMSFa (protease inhibitor) in isopropanol. The incubated cell suspension was then diluted to 2 liters total volume with demineralized water and stirred vigorously for a further 90 min. This resulted in complete lysis of the cells. To reduce the viscosity, 50 mg of DNAse was added directly after the dilution. Cell debris were removed by centrifugation (40 min at 48,OOOg, 4’C). All the following procedures took place at 4’C. The supernatant was loaded onto a DEAE-Sepharose (fast-flow) column with a bed volume of 400 ml preequilibrated with the above-mentioned phosphate buffer. The column was eluted with a 0 to 500 mM NaCl gradient (4 liters) in phosphate buffer. Greenish fractions coming off at ca. 60 mM NaCl were pooled, including several fractions eluting behind these. This pool was concentrated to 100 ml via ultrafiltration using an Amicon YM-10 membrane. The concentrated pool was diluted to 300 ml with demineralized water. CuSOl was then added to a total concentration of 0.1 mM. After 1 h of stirring ferricyanide was dissolved to a concentration of 0.2 mM. This solution of protein was applied to a 50 ml DEAE-Sepharose column as described above. The amicyanin was eluted with 50 mM phosphate buffer (pH 7.5). Care was taken to also collect fractions coming off the column after the blue fraction. These fractions still contain reduced (colorless) amicyanin. This is easily checked by adding small amounts of ferricyanide. After pooling the fractions and concentrating the solution, ferricyanide was again added in excess to amicyanin and the column pro-
3 Abbreviations used: PMSF, phenylmethylsulfonyl N-acetyltryptophanamide; Gd, guanidinium.
112 All
Copyright 0 1991 rights of reproduction
fluoride;
NATA,
0003.9861/91 $3.00 by Academic Press, Inc. in any form reserved.
INTRINSIC
FLUORESCENCE
cedure was repeated until no oxidation of eluted amicyanin was observed by ferricyanide. These columns are necessary in view of the next purification step in which the protein solution is acidified with acetic acid to pH 4.5. It is known that the reduced amicyanin is not stable under these conditions (7). The rest of the isolation procedure has been described earlier (1, 7). The yield with the present procedure amounts to 0.2 mg of pure amicyanin per gram of wet bacterial paste. The ratio of the optical absorbances at 280 and 596 nm of the purified proteins amounted to 4.20, which indicates that the protein is pure (1, 7).
OF
? ? .2 m I
Preparation of copper free amicyanin. The copper free protein was prepared by 24 h dialysis of an amicyanin sample (A,, = 1.83) vs 50 mM potassium phosphate buffer (pH 7.2) to which 10 mM KCN had been added. Cyanide was then removed by dialysis against 50 mM pbosphate buffer (pH 7.2). Cu (or Zn) could be added as their chloride or sulfates to the colorless apoprotein. The addition of C!uS04 was accompanied by the return of the blue color. Maximum color development was attained upon addition of 1 mol of CuS04 per mole of apoprotein. No significant improvement in the spectroscopic properties or in the ability of metal binding was observed if the protein was reduced with ascorbate before dialysis against KCN 1 mM. Fluorescence measurements. Fluorescence spectra were recorded with a Jobin Yvon 3D spectrofluorometer. The excitation and emission spectra were corrected by using rhodamine as a quantum counter and the reference compounds N-acetyltryptophanamide, (NATA) or p-terphenyl, respectively (8). The decay of the fluorescence was studied with the correlated single photon counting technique, using an apparatus assembled with commercial components and described elsewhere (9). The and the fluorescence was samples were excited at 295 nm (OD -0.1) collected after a 305.nm cutoff filter to remove the scattered light. The decay data were analyzed using an iterative reconvolution procedure written in Fortran and running on a PDP 11-23 Digital computer to correct for the finite width of the excitation pulse (10). The data were fitted with a sum of discrete exponential functions, f(t)
= Cioliexp(-t/r;),
where 7, are the lifetimes and 01, the preexponential factors. The adequacy of the fits was evaluated on the basis of the reduced chi-square (X2) values and of the distribution of the weighted residuals. The samples were kept at 20°C during the static and dynamic Auorescence measurements by circulating water in the sample holders. The temperature was measured directly in the cuvette.
113
AMICYANIN
50-j
: II II II ‘I II I I I 1
I :
40-
G =
30-
;‘,, . .
.I.. I
:
\
: \ \
: \ \
20-
\
,.: lo-
\
\ i’,
0 250
,:y.,
k:,
,,,, 270
290
310
330
wavelength
350
\
\
:
: \
\
370
: :.. .,.
\ 390
410
(nm)
FIG. 1. Corrected excitation and emission spectra of holoamicyanin in 50 mM phosphate buffer, pH 7.3 (-), of holoazurin in 50 mM sodium acetate buffer, pH 5 (- - -), and of holoamicyanin denatured with 6 M GdHCl (. . . ). Experimental conditions: emission wavelengths 318 f 2 nm (holoamicyanin) and 308 t 2 nm (holoaxurin), excitation wavelength 280 f 2 nm, absorbance at 280 nm - 0.1 for each protein. The spectrum of holoazurin has been amplified by 1.6 times with respect to amicyanin.
nearly to the original value (q.y. = 0.15), showing that a small fraction of the apoprotein molecules were not completely reconstituted to the holo form. The quenching effect on the apoprotein form was also studied using Zn2+. In this case a lower quenching efficiency was observed ((3.Y.
-
0.45).
Fluorescence Decay of Amicyanin
The intrinsic fluorescence of amicyanin decays in a clearly nonexponential way. A good fit of the decay profile requires a two-exponential function (Fig. 2). The lifetime RESULTS decay of holoamicyanin at pH 7.0 and 25°C can be satisfactorily fitted by two lifetimes of 0.7 and 5.7 ns, reStatic Fluorescence of Amicyanin spectively, the shorter of which accounts for about 90% Amicyanin has a slightly structured absorption spec- of the emitting species.Besides the quantum yield, copper trum in the near uv with a peak at 280 nm. Upon exci- strongly affects the decay profile. In fact in the apoprotein tation at this wavelength an emission with its peak at it becomes monoexponential with a lifetime of 3.3 ns. The 318 nm was observed. The emission spectrum shows a Zn-treated apoazurin also showed a two-lifetime emission clear fine structure (Fig. 1). The quantum yield for ho- though the values of 7 were significantly different from loamicyanin is 0.11. The excitation spectrum was the same those of apoprotein and those of Cu-containing amicyfor emission at 305,315, and 325 nm and almost identical anin. It is interesting to note that the ratio of the average to the absorption spectrum. No difference in the emission lifetimes for apo- and holoamicyanin is lower than the spectrum was observed upon changing the excitation from ratio of the respective quantum yields (2.8 vs 6.6) (Table 280 to 295 nm, ruling out an emission from the five tyI), probably due to static quenching of the holoprotein rosines present in the protein. Upon denaturation with 6 fluorescence. M guanidinium-hydrochloride (GdHCl) the emission Holo- and apoamicyanin denatured with GdHCl show spectrum shifted to 352 nm (Fig. 1). The removal of copper the same heterogeneous fluorescence decay that can be increased the quantum yield by 6.6 times without any fitted by a two-exponential function. This decay now is change in the shape of the spectrum. The addition of Cu2+ nearly identical to that of NATA under the same experto the apoprotein brought back the fluorescence intensity imental conditions.
114
ROSATO
ET
AL.
TABLE I
FluorescenceDecay Parametersof Amicyanin (AMI) at Sample 0.7
Holo-AM1 Apo-AM1 Zn-AM1 Holo-AMI + 6 M GdHCl Apo-AM1 + 6 M GdHCl NATA NATA + 6 M GdHCl
t (ns)
FIG. 2. Fluorescence decay of holoamicyanin (dots and line) and excitation profile (dots only). The solid line represents the best fit obtained with two lifetimes (cfr. Table I). In the lower part the weighted residuals of the fit are reported. Excitation wavelength = 295 zk 2 nm, emission > 305 nm (cut-off filter), absorbance at 295 nm = 0.1.
Temperature-Dependence of Amicyanin Fluorescence Decay The dependence of holoamicyanin fluorescence decay on temperature has been also studied (Fig. 3). The mean fluorescence lifetime (T(ave) = (~~7~+ (~~7~)decreases as the temperature increases from 10 to 50°C. This is mostly due to the decrease of the long component (r2) as the change of the short component (7J is in the range of experimental error. Also in the case of apoamicyanin the temperature has a strong effect on the fluorescence lifetime (Fig. 3). No hysteresis was observed in these experiments as, for instance, the measurements made at 20°C gave exactly the same results even for samples previously brought to 50°C (highest temperature reached).
Note. TV, T* = lifetimes = preexponential factor = reduced chi-square.
20°C
T (ave)
X2
1.2 2.4 1.8
0.90
0.5
5.7 3.3 4.1
0.92
1.2 3.3 0.8
2.3
6.0
0.78
3.1
3.0
2.1
6.1 3.1
0.83
2.8 3.1
1.4 2.2
2.5
5.8
0.89
2.9
2.3
in ns (67 = 0.1 ns); ~(ave) = a171 + ~~7~; o(~ relative to 71 (01* = 1 - q, bol = 0.05); X2
This finding might indicate that also in amicyanin the tryptophanyl residue is buried inside the protein and almost excluded from contact with the solvent. Furthermore it might also be inferred that copper and tryptophan are not far from each other in the three-dimensional structure of the protein, as found for azurin (12). This inference is in agreement with an NMR estimate distance of about 10 A of the tryptophan from the copper ion (G. W. Canters et al., unpublished data). Another interesting feature that amicyanin shares with azurin is the heterogeneous tluorescence decay induced by the presence of copper. The heterogeneity in azurin has been explained by the presence of at least two different protein conformations (4, 5). Petrich et al. (6) hypothesized that one of the two molecular forms is the metal-free azurin or an apo-like form. This hypothesis was made on the basis of the identity between the longer lifetime of holoazurin and the lifetime of copper-free azurin. A contaminating fluorescent speciescannot explain the fluorescence decay of amicyanin because
7 ho10 a2 10%
10%
10
20
9%
9%
8%
DISCUSSION
The intrinsic fluorescence of amicyanin shows interesting features recalling those of P. aeruginosa and P. fluorescens azurins (4, 6, 11). In fact, similarly to azurin, the fluorescence of the single tryptophanyl residue of this protein is strongly affected by the presence of copper (Table I), which reduces the quantum yield to about i with respect to the metal-free protein without effect on the shape of the emission spectrum. The emission spectrum of amicyanin at 20°C is red-shifted by about 10 nm with respect to azurin but still shows a fine structure (Fig. 1).
0
30
40
13
T ( C)
FIG. 3. perature conditions
Dependence of lifetimes and preexponential for holoamicyanin (0) and apoamicyanin as in Fig. 2.
(0).
factors on temExperimental
INTRINSIC
FLUORESCENCE
the longer lifetime of the copper-containing form (5.7 ns) is longer than that of the copper-free amicyanin (3.3 ns). The long lifetime component also cannot be ascribed to a few percent of zinc amicyanin detected in EXAFS experiments (G. W. Canters et al., unpublished data) as the decay of Zn-amicyanin was always heterogeneous and never as long as that of Cu-amicyanin (Table I). The long lifetime of the holoamicyanin is considerably shortened by increasing the temperature, accounting for almost all the quenching effect. The short lifetime of holoamicyanin is instead slightly affected by a temperature change in the range lo-50°C while the a! factors also change. No univocal explanation is available for the different behavior of the two lifetimes as a function of temperature, but certainly a simple equilibrium between two forms seems to be ruled out. In conclusion in this paper we show that many properties of amicyanin fluorescence are very similar to those of azurin. Amicyanin appears to be next to azurin in the scale of protein with anomalous blue-shifted fluorescence properties. Small but significant qualitative and quantitative differences between the two proteins may give the clue for the understanding of such properties. The heterogeneous fluorescence decay of amicyanin does not originate from a contaminating copper-free fraction. This finding should be taken into account when one discusses the fluorescence properties of other proteins like azurin.
OF
115
AMICYANIN
ACKNOWLEDGMENTS The authors thank Prof. J. A. Duine and co-workers (Delft University, The Netherlands) for their kind gift of bacterial
Technical paste.
REFERENCES 1. van Houwelingen, T., Canters, G. W., Stobbelaar, Frank, J., and Tsugita, A. (1985) Eur. J. Biochem. 2. Lontie, R. (1984) Boca Raton, FL. 3. Ambler,
Copper
Proteins
R. P., and Brown,
L. H. (1967)
4. Szabo, A. G., Stepanik, T. M., (1983)Biophys.J. 41,233-244. 5. Hutnick, 3934.
C. H., and Szabo,
6. Petrich, chemistry
J. W., Longworth, 26, 2711-2722.
7. Lommen, Biochem.
176,213-223.
8. Roberts, (Miller,
A., Canters,
D. M.,
J.
CRC Press,
104,784-825.
and Young,
Biochemistry
J. W., and Fleming,
G. W., and van Beeumen,
N. M.
28,
3923-
G. R. (1987) J. (1988)
Bio-
Eur.
J.
G. C. K. (1981) in Standards in Fluorescence Spectrometry J. N., Ed.), pp. 49-67. Chapman & Hall Ltd., London.
A., and Steinberg,
N.,
and Finazzi-Agro,
J. Z. (1974)
Anal.
11. Finazzi-Agro, A., Giovagnoli, C., Avigliano, Mondovi, B. (1973) Eur. J. Biochem. 34,20-24. 12. Adman, (1978)J.
Enzyme,
Biochem.
Wayner,
A. G. (1989)
9. Cannella, C., Berni, R., Rosato, Biochemistry 25,7319-7322. 10. Grinvald, 598.
and Copper
G., Duine, J. A., 153, 75-80.
E. T., Stenkamp, Mol. Biol. 123,
R. E., Sieker,
35-47.
Biochem.
A. (1986) 58,
L., Rotilio,
L. C., and Jensen,
583-
G., and L. H.
Vol.
OF BIOCHEMISTRY
284, No. 1, January,
AND
BIOPHYSICS
pp. 112-115,
1991
Intrinsic Fluorescence of the Bacterial CopperContaining Protein Amicyanin’ Nicola Rosato, Giampiero Mei, Isabella Savini, Arjen Lommen,* and Gerard W. Canters*
Franc0
Del Bolgia,
Dipartimento di Medicina Sperimentale & Scienze Biochimiche, Universitci *Gorlaeus Laboratories, Department of Chemistry, State University Leiden,
Received
July
Alessandro
di Roma “Tor Leiden.
Finazzi-Agrb,2
Vergata;”
and
9, 1990
The fluorescence properties of the single tryptophanyl residue present in amicyanin from Thiobacillus versutus are very similar to those of azurin from Pseudomonas aeruginosa and other mononuclear blue copper proteins. The emission maximum is well structured and centered at 3 18 nm. The quantum yield is strongly affected by the presence of copper, the removal of which is accompanied by a more than sixfold increase in fluorescence, without change in shape. The fluorescence decay of holo-amicyanin is heterogeneous with a longer component of 5.7 ns and a shorter one of 0.7 ns accounting for 90% of the total emitting molecules. Copper-free amicyanin shows instead a single exponential decay (3.3 ns) of intrinsic fluorescence. This lifetime decreases as the temperature increases as does the longer lifetime component of holoamicyanin. 0 1991 Academic Press. Inc.
Amicyanin is a member of the family of bacterial and plant mononuclear blue copper proteins (1). Its intense blue color, EPR spectrum, redox potential, copper content, molecular weight, and amino acid composition recall those of azurin, plastocyanin, plantacyanin, rusticyanin, and so on (2). The presence of a single tryptophan in amicyanin analogous to Pseudomonas aeruginosa and Pseudomonas fluorescens azurins (3) suggested to us a study of its static and dynamic fluorescence. The fluorescence of azurins often shows very peculiar spectroscopic properties, becausethe single tryptophan shows a very complicated fluorescence decay. The origin of this heterogeneous decay is still a matter of controversy (4, 6). 1 This work has been partly supported by Italian Minister0 della Pubblica Istruzione “Progetti Nazionali 40%.” ‘To whom correspondence should be addressed at Dipartimento di Medicina Sperimentale e Scienze Biochimiche, Universith di Roma “Tor Vergata,” Via 0. Raimondo, 00173 Roma, Italy; 6-7230291; Fax: 6-6130793.
In the present paper we report that amicyanin also has static and dynamic fluorescence properties very different from those of simple tryptophan model compounds but similar to those of the bacterial protein azurin. The results presented here rule out some hypotheses on the complex decay in single tryptophan-containing proteins. MATERIALS
AND
METHODS
Purification of amicyanin. The bacterial paste of Thiobacillus uersutus ATCC 25364T (l), which was used for the preparation of amicyanin, was taken from a supply grown previously (7). The isolation procedure for amicyanin was modified as follows: 200 g of wet bacterial paste was resuspended in 300 ml of IO mM phosphate buffer (pH 7.0). The cell suspension was incubated for 1 h at 15’C with 200 mg lysozyme, 180 mg EDTA, and 0.5 ml of a solution of 100 mM PMSFa (protease inhibitor) in isopropanol. The incubated cell suspension was then diluted to 2 liters total volume with demineralized water and stirred vigorously for a further 90 min. This resulted in complete lysis of the cells. To reduce the viscosity, 50 mg of DNAse was added directly after the dilution. Cell debris were removed by centrifugation (40 min at 48,OOOg, 4’C). All the following procedures took place at 4’C. The supernatant was loaded onto a DEAE-Sepharose (fast-flow) column with a bed volume of 400 ml preequilibrated with the above-mentioned phosphate buffer. The column was eluted with a 0 to 500 mM NaCl gradient (4 liters) in phosphate buffer. Greenish fractions coming off at ca. 60 mM NaCl were pooled, including several fractions eluting behind these. This pool was concentrated to 100 ml via ultrafiltration using an Amicon YM-10 membrane. The concentrated pool was diluted to 300 ml with demineralized water. CuSOl was then added to a total concentration of 0.1 mM. After 1 h of stirring ferricyanide was dissolved to a concentration of 0.2 mM. This solution of protein was applied to a 50 ml DEAE-Sepharose column as described above. The amicyanin was eluted with 50 mM phosphate buffer (pH 7.5). Care was taken to also collect fractions coming off the column after the blue fraction. These fractions still contain reduced (colorless) amicyanin. This is easily checked by adding small amounts of ferricyanide. After pooling the fractions and concentrating the solution, ferricyanide was again added in excess to amicyanin and the column pro-
3 Abbreviations used: PMSF, phenylmethylsulfonyl N-acetyltryptophanamide; Gd, guanidinium.
112 All
Copyright 0 1991 rights of reproduction
fluoride;
NATA,
0003.9861/91 $3.00 by Academic Press, Inc. in any form reserved.
INTRINSIC
FLUORESCENCE
cedure was repeated until no oxidation of eluted amicyanin was observed by ferricyanide. These columns are necessary in view of the next purification step in which the protein solution is acidified with acetic acid to pH 4.5. It is known that the reduced amicyanin is not stable under these conditions (7). The rest of the isolation procedure has been described earlier (1, 7). The yield with the present procedure amounts to 0.2 mg of pure amicyanin per gram of wet bacterial paste. The ratio of the optical absorbances at 280 and 596 nm of the purified proteins amounted to 4.20, which indicates that the protein is pure (1, 7).
OF
? ? .2 m I
Preparation of copper free amicyanin. The copper free protein was prepared by 24 h dialysis of an amicyanin sample (A,, = 1.83) vs 50 mM potassium phosphate buffer (pH 7.2) to which 10 mM KCN had been added. Cyanide was then removed by dialysis against 50 mM pbosphate buffer (pH 7.2). Cu (or Zn) could be added as their chloride or sulfates to the colorless apoprotein. The addition of C!uS04 was accompanied by the return of the blue color. Maximum color development was attained upon addition of 1 mol of CuS04 per mole of apoprotein. No significant improvement in the spectroscopic properties or in the ability of metal binding was observed if the protein was reduced with ascorbate before dialysis against KCN 1 mM. Fluorescence measurements. Fluorescence spectra were recorded with a Jobin Yvon 3D spectrofluorometer. The excitation and emission spectra were corrected by using rhodamine as a quantum counter and the reference compounds N-acetyltryptophanamide, (NATA) or p-terphenyl, respectively (8). The decay of the fluorescence was studied with the correlated single photon counting technique, using an apparatus assembled with commercial components and described elsewhere (9). The and the fluorescence was samples were excited at 295 nm (OD -0.1) collected after a 305.nm cutoff filter to remove the scattered light. The decay data were analyzed using an iterative reconvolution procedure written in Fortran and running on a PDP 11-23 Digital computer to correct for the finite width of the excitation pulse (10). The data were fitted with a sum of discrete exponential functions, f(t)
= Cioliexp(-t/r;),
where 7, are the lifetimes and 01, the preexponential factors. The adequacy of the fits was evaluated on the basis of the reduced chi-square (X2) values and of the distribution of the weighted residuals. The samples were kept at 20°C during the static and dynamic Auorescence measurements by circulating water in the sample holders. The temperature was measured directly in the cuvette.
113
AMICYANIN
50-j
: II II II ‘I II I I I 1
I :
40-
G =
30-
;‘,, . .
.I.. I
:
\
: \ \
: \ \
20-
\
,.: lo-
\
\ i’,
0 250
,:y.,
k:,
,,,, 270
290
310
330
wavelength
350
\
\
:
: \
\
370
: :.. .,.
\ 390
410
(nm)
FIG. 1. Corrected excitation and emission spectra of holoamicyanin in 50 mM phosphate buffer, pH 7.3 (-), of holoazurin in 50 mM sodium acetate buffer, pH 5 (- - -), and of holoamicyanin denatured with 6 M GdHCl (. . . ). Experimental conditions: emission wavelengths 318 f 2 nm (holoamicyanin) and 308 t 2 nm (holoaxurin), excitation wavelength 280 f 2 nm, absorbance at 280 nm - 0.1 for each protein. The spectrum of holoazurin has been amplified by 1.6 times with respect to amicyanin.
nearly to the original value (q.y. = 0.15), showing that a small fraction of the apoprotein molecules were not completely reconstituted to the holo form. The quenching effect on the apoprotein form was also studied using Zn2+. In this case a lower quenching efficiency was observed ((3.Y.
-
0.45).
Fluorescence Decay of Amicyanin
The intrinsic fluorescence of amicyanin decays in a clearly nonexponential way. A good fit of the decay profile requires a two-exponential function (Fig. 2). The lifetime RESULTS decay of holoamicyanin at pH 7.0 and 25°C can be satisfactorily fitted by two lifetimes of 0.7 and 5.7 ns, reStatic Fluorescence of Amicyanin spectively, the shorter of which accounts for about 90% Amicyanin has a slightly structured absorption spec- of the emitting species.Besides the quantum yield, copper trum in the near uv with a peak at 280 nm. Upon exci- strongly affects the decay profile. In fact in the apoprotein tation at this wavelength an emission with its peak at it becomes monoexponential with a lifetime of 3.3 ns. The 318 nm was observed. The emission spectrum shows a Zn-treated apoazurin also showed a two-lifetime emission clear fine structure (Fig. 1). The quantum yield for ho- though the values of 7 were significantly different from loamicyanin is 0.11. The excitation spectrum was the same those of apoprotein and those of Cu-containing amicyfor emission at 305,315, and 325 nm and almost identical anin. It is interesting to note that the ratio of the average to the absorption spectrum. No difference in the emission lifetimes for apo- and holoamicyanin is lower than the spectrum was observed upon changing the excitation from ratio of the respective quantum yields (2.8 vs 6.6) (Table 280 to 295 nm, ruling out an emission from the five tyI), probably due to static quenching of the holoprotein rosines present in the protein. Upon denaturation with 6 fluorescence. M guanidinium-hydrochloride (GdHCl) the emission Holo- and apoamicyanin denatured with GdHCl show spectrum shifted to 352 nm (Fig. 1). The removal of copper the same heterogeneous fluorescence decay that can be increased the quantum yield by 6.6 times without any fitted by a two-exponential function. This decay now is change in the shape of the spectrum. The addition of Cu2+ nearly identical to that of NATA under the same experto the apoprotein brought back the fluorescence intensity imental conditions.
114
ROSATO
ET
AL.
TABLE I
FluorescenceDecay Parametersof Amicyanin (AMI) at Sample 0.7
Holo-AM1 Apo-AM1 Zn-AM1 Holo-AMI + 6 M GdHCl Apo-AM1 + 6 M GdHCl NATA NATA + 6 M GdHCl
t (ns)
FIG. 2. Fluorescence decay of holoamicyanin (dots and line) and excitation profile (dots only). The solid line represents the best fit obtained with two lifetimes (cfr. Table I). In the lower part the weighted residuals of the fit are reported. Excitation wavelength = 295 zk 2 nm, emission > 305 nm (cut-off filter), absorbance at 295 nm = 0.1.
Temperature-Dependence of Amicyanin Fluorescence Decay The dependence of holoamicyanin fluorescence decay on temperature has been also studied (Fig. 3). The mean fluorescence lifetime (T(ave) = (~~7~+ (~~7~)decreases as the temperature increases from 10 to 50°C. This is mostly due to the decrease of the long component (r2) as the change of the short component (7J is in the range of experimental error. Also in the case of apoamicyanin the temperature has a strong effect on the fluorescence lifetime (Fig. 3). No hysteresis was observed in these experiments as, for instance, the measurements made at 20°C gave exactly the same results even for samples previously brought to 50°C (highest temperature reached).
Note. TV, T* = lifetimes = preexponential factor = reduced chi-square.
20°C
T (ave)
X2
1.2 2.4 1.8
0.90
0.5
5.7 3.3 4.1
0.92
1.2 3.3 0.8
2.3
6.0
0.78
3.1
3.0
2.1
6.1 3.1
0.83
2.8 3.1
1.4 2.2
2.5
5.8
0.89
2.9
2.3
in ns (67 = 0.1 ns); ~(ave) = a171 + ~~7~; o(~ relative to 71 (01* = 1 - q, bol = 0.05); X2
This finding might indicate that also in amicyanin the tryptophanyl residue is buried inside the protein and almost excluded from contact with the solvent. Furthermore it might also be inferred that copper and tryptophan are not far from each other in the three-dimensional structure of the protein, as found for azurin (12). This inference is in agreement with an NMR estimate distance of about 10 A of the tryptophan from the copper ion (G. W. Canters et al., unpublished data). Another interesting feature that amicyanin shares with azurin is the heterogeneous tluorescence decay induced by the presence of copper. The heterogeneity in azurin has been explained by the presence of at least two different protein conformations (4, 5). Petrich et al. (6) hypothesized that one of the two molecular forms is the metal-free azurin or an apo-like form. This hypothesis was made on the basis of the identity between the longer lifetime of holoazurin and the lifetime of copper-free azurin. A contaminating fluorescent speciescannot explain the fluorescence decay of amicyanin because
7 ho10 a2 10%
10%
10
20
9%
9%
8%
DISCUSSION
The intrinsic fluorescence of amicyanin shows interesting features recalling those of P. aeruginosa and P. fluorescens azurins (4, 6, 11). In fact, similarly to azurin, the fluorescence of the single tryptophanyl residue of this protein is strongly affected by the presence of copper (Table I), which reduces the quantum yield to about i with respect to the metal-free protein without effect on the shape of the emission spectrum. The emission spectrum of amicyanin at 20°C is red-shifted by about 10 nm with respect to azurin but still shows a fine structure (Fig. 1).
0
30
40
13
T ( C)
FIG. 3. perature conditions
Dependence of lifetimes and preexponential for holoamicyanin (0) and apoamicyanin as in Fig. 2.
(0).
factors on temExperimental
INTRINSIC
FLUORESCENCE
the longer lifetime of the copper-containing form (5.7 ns) is longer than that of the copper-free amicyanin (3.3 ns). The long lifetime component also cannot be ascribed to a few percent of zinc amicyanin detected in EXAFS experiments (G. W. Canters et al., unpublished data) as the decay of Zn-amicyanin was always heterogeneous and never as long as that of Cu-amicyanin (Table I). The long lifetime of the holoamicyanin is considerably shortened by increasing the temperature, accounting for almost all the quenching effect. The short lifetime of holoamicyanin is instead slightly affected by a temperature change in the range lo-50°C while the a! factors also change. No univocal explanation is available for the different behavior of the two lifetimes as a function of temperature, but certainly a simple equilibrium between two forms seems to be ruled out. In conclusion in this paper we show that many properties of amicyanin fluorescence are very similar to those of azurin. Amicyanin appears to be next to azurin in the scale of protein with anomalous blue-shifted fluorescence properties. Small but significant qualitative and quantitative differences between the two proteins may give the clue for the understanding of such properties. The heterogeneous fluorescence decay of amicyanin does not originate from a contaminating copper-free fraction. This finding should be taken into account when one discusses the fluorescence properties of other proteins like azurin.
OF
115
AMICYANIN
ACKNOWLEDGMENTS The authors thank Prof. J. A. Duine and co-workers (Delft University, The Netherlands) for their kind gift of bacterial
Technical paste.
REFERENCES 1. van Houwelingen, T., Canters, G. W., Stobbelaar, Frank, J., and Tsugita, A. (1985) Eur. J. Biochem. 2. Lontie, R. (1984) Boca Raton, FL. 3. Ambler,
Copper
Proteins
R. P., and Brown,
L. H. (1967)
4. Szabo, A. G., Stepanik, T. M., (1983)Biophys.J. 41,233-244. 5. Hutnick, 3934.
C. H., and Szabo,
6. Petrich, chemistry
J. W., Longworth, 26, 2711-2722.
7. Lommen, Biochem.
176,213-223.
8. Roberts, (Miller,
A., Canters,
D. M.,
J.
CRC Press,
104,784-825.
and Young,
Biochemistry
J. W., and Fleming,
G. W., and van Beeumen,
N. M.
28,
3923-
G. R. (1987) J. (1988)
Bio-
Eur.
J.
G. C. K. (1981) in Standards in Fluorescence Spectrometry J. N., Ed.), pp. 49-67. Chapman & Hall Ltd., London.
A., and Steinberg,
N.,
and Finazzi-Agro,
J. Z. (1974)
Anal.
11. Finazzi-Agro, A., Giovagnoli, C., Avigliano, Mondovi, B. (1973) Eur. J. Biochem. 34,20-24. 12. Adman, (1978)J.
Enzyme,
Biochem.
Wayner,
A. G. (1989)
9. Cannella, C., Berni, R., Rosato, Biochemistry 25,7319-7322. 10. Grinvald, 598.
and Copper
G., Duine, J. A., 153, 75-80.
E. T., Stenkamp, Mol. Biol. 123,
R. E., Sieker,
35-47.
Biochem.
A. (1986) 58,
L., Rotilio,
L. C., and Jensen,
583-
G., and L. H.
