Purification, crystallization, NMR spectroscopy and biochemical analyses of a-phycoerythrocyanin peptides Georg Wiegand 1 , Axel Parbel 2 *, Markus H. J. Seifert 1 , Tad A. Holak 1 and Wolfgang Reuter 1 1 Max-Planck-Institut fu ¨ r Biochemie, Martinsried, Germany; 2 Botanisches Institut der Ludwig-Maximilians-Universita ¨ t, Mu ¨ nchen, Germany The a-phycoerythrocyanin subunits of the different phy- coerythrocyanin complexes of the phycobilisomes from the cyanobacterium Mastigocladus laminosus perform a remarkable photochemistry. Similar to phytochromes – the photoreceptors of higher plants – the spectral pro- perties of the molecule reversibly change according to the irradiation wavelength. To enable extensive analyses, the protein has been produced at high yield by improving purification protocols. As a result, several comparative studies on the Z-andE-configurations of the intact a-subunit, and also on photoactive peptides originating from nonspecific degradations of the chromoprotein, were possible. The analyses comprise absorbance, fluorescence and CD spectroscopy, crystallization, preliminary X-ray measurements, mass spectrometry, N-terminal amino acid sequencing and 1D NMR spectroscopy. Intact a-phyco- erythrocyanin aggregates significantly, due to hydrophobic interactions between the two N-terminal helices. Removal of these helices reduces the aggregation but also desta- bilizes the protein fold. The complete subunit could be crystallized in its E-configuration, but the X-ray meas- urement conditions must be improved. Nevertheless, NMR spectroscopy on a soluble photoactive peptide presents the first insight into the complex chromophore protein interactions that are dependent on the light induced state. The chromophore environment in the Z-configuration is rigid whereas other regions of the protein are more flexible. In contrast, the E-configuration has a mobile chromophore, especially the pyrrole ring D, while other regions of the protein rigidified compared to the Z-configuration. Keywords: phycobilisomes; phycoerythrocyanin; protein structure; photochemistry; energy transfer. Phycobiliproteins are a class of chromoproteins bearing covalently bound linear tetrapyrrole (phycobilins) chro- mophores. They are predominantly involved in the photo- synthetic light harvesting process of cyanobacteria and certain algae. With respect to this function they are assembled to supramolecular protein pigment complexes, i.e. phycobilisomes, building up a highly ordered network of very rigid chromophores which enable an energy transfer efficiency of almost 100% [1,2]. The phycoerythrocyanin (PEC) complexes located at the periphery of phycobilisomes are present in only a few species of cyanobacteria [3]. PEC of the thermophilic cyanobacterium Mastigocladus laminosus is the best char- acterized complex of this biliprotein class. Like other peripheral phycobiliproteins, e.g. phycoerythrin, low light and high temperature conditions induce a maximum content of PEC, reaching approximately 30% of total protein content within the phycobilisomes [4,5]. The X-ray structure of PEC reveals three heterodimeric a,b substruc- tures, so called ‘monomers’, which are associated to a ring shaped disc, designated as ‘trimer’. The b-subunit contains two phycocyanobilin (PCB) chromophores, whereas the a-subunit has a single phycoviolobilin (PVB) chromophore of which the pyrrole ring A is covalently linked via a thioether bond to Cys84 [6]. Unlike other phycobiliproteins, the PVB chromophores of the PEC-‘trimers’ present a remarkable reversible photochemistry which has been reported first by Bjo ¨ rn [7]. The observation initiated intense investigations into this unusual spectroscopic behavior, especially regarding the possible function as a sensor pigment [8–10]. The sensor function seems to be questionable, however, in particular because of the high content and extremely reduced photochemistry of PEC within the phycobilisomes [4,11,12]. Biochemical and spectral data assign the photo- chemistry exclusively to the a-PEC subunit [13,14]. Similar to phytochrome and phytochrome-like photoreceptors of higher plants and cyanobacteria, the PVB chromophore undergoes spectral and molecular changes depending on light quality [15,16]. The phototransformation of a-PEC is reflected by the reversible shift in the visible absorption maximum from 505 to 570 nm and the two states were termed E and Z, respectively. Isomerization can be performed by irradiation with complementary chromatic light and the two states are quite stable in the dark. The molecular mechanism of the photoreaction is similar to that of the phytochromes and is proposed to exists as a chromophore twisting around the D15,16 double bond between the C and D pyrrole rings [9,16]. However, the Correspondence to W. Reuter, Max-Planck-Institut fu ¨ rBiochemie, Am Klopferspitz 18 A, 82152 Martinsried, Germany. Fax: + 49 (0)89 85783516, Tel.: + 49 (0)89 85782707, E-mail: reuter@biochem.mpg.de Abbreviations: FID, free induction decay; HIC, hydrophobic interac- tion chromatography; MPD, 2-methyl-2,4-pentanediol; PCB, phyco- cyanobilin; PEC, phycoerythrocyanin; PVB, phycoviolobilin. *Present address: Amersham Biosciences, Munzinger Str. 9, 79111 Freiburg, Germany. (Received 4 June 2002, revised 28 August 2002, accepted 29 August 2002) Eur. J. Biochem. 269, 5046–5055 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03221.x situation may be more complicated in a-PEC, as at least type I and type II Z/E-isomerizations have been spectrosc- opically discerned. The ratio of the two types has been suggested to be controlled by sulfhydryls of Cys98 and Cys99 in the protein [8,10]. Whereas, the structures of the PVB isomers are well defined by spectroscopical charac- terizations, the participitation of the protein in the phototransformation process of the a-subunit is almost unknown. Some reasons for this may be the considerable difficulties in the preparative purification of a-PEC and the problems arising during analyses of the protein at high concentrations. The present study describes a very effective method of isolation and purification of PEC and its photoactive a-subunit. Crystallization and preliminary X-ray experi- ments suggest pronounced conformational alterations of both the protein and the PVB chromophore, depending on light quality. N-Terminal amino acid analyses and mass spectrometry characterized a series of a-PEC peptides which occurred during storage in formic acid. The Z-and E-configurations of one chromopeptide with an excellent solubility and stability were investigated by 1D NMR spectroscopy. The information presented here are mainly focused on the preparations, handling and characterizations of a-PEC peptides which will be a prerequisite for successful studies on the detailed molecular events during phototrans- formation. MATERIALS AND METHODS Strain, culturing conditions and isolation of phycobilisomes The thermophilic cyanobacterium M. laminosus (syn. Fischerella sp. PCC 7603) was photoautotropically grown at 48 °C, 10 lEÆm )2 Æs )1 andgassedwith2%(v/v)CO 2 enriched air. The cells were harvested when an optical density of 0.7 at 740 nm was reached. At these conditions the phycobilisomes are attributed by the maximum content of PEC [4]. The phycobilisomes were isolated by step-gradient cen- trifugation principally following the buffer conditions described by Reuter and Wehrmeyer [17]. Scaling up of the phycobilisome preparation was necessary for the development of the effective purification procedure of the a-subunit. Cells of M. laminosus with a wet weight of 40 g were disrupted at 17 °C in a self-constructed cell mill. The glass beads (1 mm) were filtered off and the filtrate of about 180 mL was incubated with 4% (w/v) N,N-dimethyldode- cylamine N-oxide (Fluka, Buchs, Switzerland) for 1 h at 17 °C. This homogenate was clarified by centrifugation at 48 000 g for 30 min at 17 °C. The resulting supernatant of approximately 180 mL was directly applied to step-gradi- ents comprising 10 mL 40% (w/v), 15 mL 30% (w/v), 15 mL 20% (w/v) and 15 mL 10% (w/v) sucrose, respect- ively. Centrifugation was performed in a Ti45 rotor (Beckman Coulter, USA) at 40 000 g for 16 h at 17 °C. The separated phycobilisomes banding between 40% (w/v) and 30% (w/v) sucrose were eluted and subsequently precipitated by a final concentration of 2.0 M potassium phosphate, pH 7.0. These products are stable at 4 °Cforat least 2 years without any changes in their protein compo- sition. Purification of PEC complexes About 1000 mg of precipitated phycobilisomes were sedi- mented by centrifugation at 74 000 g for 30 min at 17 °C. The sediment was resolved in distilled water and dissoci- ation was performed by gel filtration on a Sephadex G25 column (Amersham Biosciences, 40 mm · 250 mm) and equilibrated with 5 m M potassium phosphate, pH 7.0. The eluted phycobilisomes were applied directly to an anion exchange column (40 mm · 150 mm) packed with DEAE- Trisacryl M (Serva, Heidelberg, Germany) and equilibrated with the dissociation buffer. At this pH and ionic strength the PEC complexes with and without linker polypeptides eluted completely, whereas all other biliprotein complexes remained on the column. The resulting PEC fraction of about 250 mg was precipitated with a final concentration of 2.0 M potassium phosphate, pH 7.0. Purification of the a-PEC subunit Sedimented PEC was resolved in distilled water and the biliprotein complexes were dissociated by gel filtration on a PD-10 column (Amersham Biosciences) in 0.3% (v/v) formic acid. The dissociation into a-andb-subunits is accompanied by the loss of the excitonic coupling between the corresponding chromophores. Therefore, the complete- ness of dissociation can be followed by the color change of PEC from pink to blue. Separation of the subunits from each other was obtained by isocratic hydrophobic interac- tion chromatography in 0.3% (v/v) formic acid on a column (10 mm · 30 mm) packed with isopropyl-substi- tuted Sephacryl-S300 (Amersham Biosciences). The proce- dure of the isopropyl substitution of the gel will be published elsewhere. During the chromatography, the a-subunits show negligible interactions with the gel, whereas the elution of b-subunits and linker polypeptides is strongly retarded. The eluted homogeneous a-PEC fraction of at least 80 mg was concentrated by ultrafiltra- tion up to 20 mgÆmL )1 . Figure 1 summarizes the purification steps which are necessary for the isolation of the intact, photoactive a-subunit. Some of the different steps are based on previous methods, e.g. induction of the maximal PEC content [4,5], anion exchange chromatography on DEAE [18], or use of formic acid as isolation medium [19]. Nevertheless, some advantages of the preparation meth- ods should be described. Based on a maximal phycobilisome concentration of 30 mgÆmL )1 within the cells [20], the isolation yield of about 80% of intact phycobilisomes is very high. The effectiveness of the cell breakage (¼ 95%) and the complete precipitation of the phycobilisomes without dis- sociation are responsible for the extraordinary high yield (results not shown). A previous study described an unusual fractionation of PEC on DEAE–cellulose [18]. This non- specific separation was not observed on DEAE-Trisacryl M, therefore the elution time and dilution of the sample was considerably reduced. In addition, short-time dissociation by gel filtration with formic acid and the fast separation by hydrophobic interaction chromatography (HIC) reduce the time consumption and the denaturation effects. How- ever, the most important step was the chromatography on the isopropyl-substituted Sephacryl-S300 minimizing non- specific interactions of a-PEC with the gel matrix. Ó FEBS 2002 Analyses of a-phycoerythrocyanin peptides (Eur. J. Biochem. 269) 5047 Optical spectroscopy The spectra of a-PEC were recorded after saturated irradiation with 577 nm light transforming the E-isomer and 500 nm light transforming the Z-isomer, respectively. Absorbance spectra were measured with a Lambda 2 UV/ visible spectrophotometer (Perkin-Elmer) and circular dichroism (CD) was recorded on a Dichrograph VI (ISA). The spectral band width was 0.25 nm, the scan speed 5nmÆs )1 . Fluorescence spectra were recorded with 2 nm resolution on a Spex model 221 fluorimeter. Details of the measurements are described by Parbel et al. [21]. Crystallization of a-PEC Unless otherwise stated, all preparations were carried out under red light with an emittance maximum of 650 nm (Phillips, the Netherlands; TLD 15). Crystallization growth was controlled in a modified microscope at a wavelength of 620 nm. Parallel crystallization experiments were conducted after transforming the a-PEC in the E-andZ-state, respectively. The state of the two isomers was monitored by absorbance spectrometry in the range of 250–650 nm. Using the vapor diffusion hanging-drop method, the proteins were crystallized in a pH range of 4.0–8.5 because at these pH values both protein and chromophore show a high stability. Crystallization of the E-form could only be observed in the presence of different salts, e.g. ammonium sulfate, ammonium phosphate, sodium-potassium phos- phate or Tris phosphate, as precipitants. Other precipitants like poly(ethylene glycol) or 2-methyl-2,4-pentanediol (MPD) have not, as yet, been found to be successful. Crystals of the Z-form have never been obtained, although the crystallization conditions of both protein states have been identical. Thin plates with dimensions of approxi- mately 0.2 · 0.6 · 0.3 mm of the E-isomer grown in the presence of 1.0 M dibasic Tris phosphate, pH 8.0, at 18 °C were analyzed by diffraction studies and mass spectrometry. For cryo-measurements the crystals were transferred into 3 M Tris phosphate, pH 8.0, which serves as cryo-protec- tant. The crystals were frozen directly in liquid nitrogen and the X-ray diffractions were recorded under white room-light at temperatures between )140 and )160 °C. ‘Native’ PAGE PAGE was performed in 3-mm thick 10% (w/v) polyacryl- amide slab gels with Tris/boric acid (42 m M /100 m M , pH 7.9). Gels were polymerized with 0.1% (v/v) tetrameth- ylethylenediamine and 0.03% (w/v) ammonium peroxodi- sulfate [17]. Samples of 1.5 mL containing 10–15 mgÆmL )1 protein were electrophoresed in the Tris/boric acid buffer system for 2400 VÆh )1 at a constant power of 18 W, at 10 °C and continuous buffer circulation in a DESAGA VA-200 apparatus (DESAGA, Germany). After separ- ation, the protein bands were cut out and the gel slices were squeezed between two glass plates. The homogeneous gel pastes were eluted for 3 h under continuous stirring with a 10-fold volume of water. After elution the homogenates were centrifuged for 1 h at 75 000 g and the supernatants were filtered through a 0.22-lm poly(vinylidene difluoride) membrane (Millipore, USA) [22]. The peptides were con- centrated by ultrafiltration and the photoactivity was tested by absorbance spectra after alternative irradiation with the two light qualities. Mass spectrometry and N-terminal amino acid sequencing Mass spectrometry of the a-PEC peptides originating from the preparation of the hydrophobic interaction column, the crystals and the ‘native’ PAGE was performed by the electrospray method in a single quadrupol mass spectro- meter API165 (Applied Biosystems, Langen, Germany). The spectra were deconvoluted with the BIOTOOL software of the manufacturer. Removal of salts within the samples was performed by hydrophobic interaction chromato- graphy on ReproSil-Pur C18-AQ, 3l,1· 150 mm (Dr A. Maisch, Ammerbuch, Germany). The peptides were eluted with a gradient from 10% (v/v) trifluoroacetic acid in H 2 O to 0.08% (v/v) trifluoroacetic acid in acetonitrile. All sequences were determined with an Applied Biosys- tems sequencer model 473 A following the manufacturer’s instructions. Structural comparison of the intact a-PEC and the degraded a-PEC peptide 2 The picture comparing the secondary structures of the intact subunit and peptide 2 was produced with the 3D Fig. 1. Overview of the isolation and purification steps of a-PEC from Mastigocladus laminosus. Approximately 250 mg of PEC-linker com- plexes could be separated from 1000 mg of phycobilisomes. The pre- cipitated PEC-fraction can be stored without alteration at least for a period of 2 years. Starting with 250 mg of the linker-PEC complexes, the purification at dissociating conditions in 0.3% (v/v) formic acid results in a final preparation of approximately 80 mg of homogeneous a-PEC. 5048 G. Wiegand et al. (Eur. J. Biochem. 269) Ó FEBS 2002 visualization program ‘WebLab ViewerPro, Version 3.20’ (Molecular Simulations Inc.). The coordinates derived from the structure analyses of phycoerythrocyanin [6]. Light-dependent 1D NMR spectroscopy of a-PEC peptide 2 Prior to NMR measurements, peptide 2 in 20 m M sodium- potassium phosphate, pH 7.2 with a protein concentration of 15–20 mgÆmL )1 was irradiated with light of 571 and 503 nm inducing the E-andZ-configuration, respectively. The complete transfer into both configurations was obtained by an illumination time of 1 h. Continuous spinning of the NMR tube minimized the self-shadowing of the highly concentrated sample. 1 H-NMR measurements were carried out in the dark without spinning on a Bruker DRX600 spectrometer equipped with a 1 H- 13 C- 15 Ntriple- resonance probehead including triple-axis gradients. All spectra were recorded at a temperature of 27 °C. To suppress the water resonance a jump-return pulse sequence was used [23]. For each spectrum 512 free induction decays (FIDs) with 32 k time domain points comprising a sweep width of 9 kHz were recorded. The interscan delay was set to 1 s. The 90° pulse was determined to be 8.4 ls. The spectra were processed by fast Fourier transformation including a Gaussian window function and digital filtering of low frequencies in the range of 1.5 p.p.m. to enhance water suppression. Only 12 k of the recorded 32 k time domain points were used for transformation to increase signal-to-noise ratio. The final spectra were processed to 32 k frequency domain points. RESULTS Spectral behavior of a-PEC in 0.3% (v/v) formic acid The steady state absorption, fluorescence and CD spectra of a-PEC depending on pre-illumination are represented in Fig. 2. The a-subunit in the E-configuration is characterized by a long wavelength absorbance maximum at 503 nm with a pronounced shoulder near 566 nm, an extremely low fluorescence and a CD minimum near 505 nm. The sharp peak (arrowhead) in the fluorescence spectrum originates from the excitation light. The absorbance shoulder near 566 nm, the broad fluorescence maximum at 588 (arrow) and the minimum in the CD spectrum near 325 nm are typical for signals of a-PEC in the Z-configuration. Therefore, the presence of these signals in the spectra of the E-isomer indicates an incomplete transformation of the molecule or at least of the chromophore. In contrast, the Z-configuration of a-PEC reveals uniform maxima at 566 nm (absorbance), 588 nm (fluorescence), 566 nm (CD) and a minimum at 329 nm (CD). The spectral data of the proteins in the E-as well as the Z-state in 0.3% (v/v) formic acid are nearly equal to those in conventional buffers near pH 7.0 [9,10,12]. Thus, the ‘native’ state of the chromoprotein has been assumed. Crystallization of a-PEC In order to obtain information about changes of the polypeptide properties in the Z-andE-state, respectively, crystallization was performed with the protein in both configurations. One major problem was the aggregation Fig. 2. Optical spectroscopy of the E- and Z-configurations of the a-subunit in 0.3% (v/v) formic acid. The spectral behavior of the chromoprotein in is nearly identical to that at neutral pH which con- firms the suitability of the isolation and purification method. It must be noted that the chromophore cannot completely be transferred into the E-configuration. This is shown by the arrows in the absorbance, fluorescence and CD spectra. The fluorescence of a-PEC in the E-configuration is extremely low, therefore the excitation light, marked by an arrowhead, is seen in the spectrum. Ó FEBS 2002 Analyses of a-phycoerythrocyanin peptides (Eur. J. Biochem. 269) 5049 behavior of the protein at pH values near 7.0, especially in the Z-configuration. Therefore, variations in the protein concentrations (5–7.5 mgÆmL )1 ) were strongly limited. The crystallization behavior of the E-isomer is identical in the dark and in green light (results not shown). This was tested by parallel crystallization attempts in the dark and under weak monochromatic green light. All common precipitants have been used but only different salts at varying ionic strength and pH values have been successful. Two typical crystallization conditions comparing the E-andthe Z-configurations are demonstrated in Fig. 3. Despite the identical crystallization conditions, only the E-configuration crystallized (Fig. 3a,b), whereas the Z-configuration always showed a type of phase separation (Fig. 3c,d). The branch- ing of the crystals occurred under nearly all conditions, however, the size of single, homogeneous crystal plates was sufficient for further analyses. Unfortunately, X-ray analyses of such plates were unsuccessful because the diffraction of the crystals decreased very rapidly during the measurements. This phenomenon has been observed for different crystals, even at low temperatures between )140 and )160 °C. Because of the extreme changes in the diffraction patterns, a unique determination of the space group and the unit cell was not possible. Nevertheless, within the limits of the measure- ments, we tentatively determined an orthorhombic space group with two molecules in the asymmetric unit. What is the reason of the strongly decreasing diffractions? The frozen crystals have been mounted and measured under white room-light. At this condition, light-induced conform- ational changes which destroy the well ordered crystal packing might be possible. The molecular flexibility of different crystallized phycobiliproteins at temperatures in the range from )100 to )160 °C has frequently been observed during the freezing and measuring procedures (Reuter, unpublished results). In addition, different inter- mediate chromophore states of PEC were recorded depend- ing on the measuring temperatures [24,25]. The results clearly demonstrate the molecular mobility of phycobili- proteins, even at low temperatures, but the influence of light on the crystal packing of a-PEC during the measurements remains uncertain. Purification and analyses of a-PEC peptides The storage time of a-PEC in 0.3% (v/v) formic acid at 4 °C was approximately 6 months. At the end of this time, the crystals shown in Fig. 3 could not be reproduced. This fact initiated an analysis of the sample by mass spectrometry revealing at least seven peptides with molecular weights between 16 000 and 14 000 Da (results not shown). At present, the reasons for the degradation are uncertain. A proteolytic splitting of the a-subunit by proteases may be possible, although the pH of 2.2 of the formic acid probably inhibits the activity of most peptidases. Another postulation is the acid-induced degradation of the a-PEC during long- term storage. Specific acid-catalyzed degradation reactions have previously been reported for other proteins [26]. The most probable explanation would be a nonspecific acid- catalyzed hydrolysis of a-PEC which is facilitated by a partial unfolding of the two N-terminal helices (Fig. 4). The resulting high flexibility of this peptide region may be responsible for destabilization of favored peptide bonds within the protein. This view is in line with the variability of the amino acid sequences for which the degradation occurs. However, cooperation between the three mechanisms cannot be excluded. Further studies on the instability of the isolated a-subunit are in progress and some aspects will be stressed in the discussion section. Fig. 3. Crystallization of a-PEC has been successful only with the molecule in the E-configuration (a,b). In principle, all crystals have been grown at 17 °C by the hanging-drop method with vapor diffusion concentration. Only salt precipitation resulted in crystals as shown in the figure. (a) Potassium phosphate, pH 7.5; (b) Tris phosphate, pH 8.0; (c) potassium phosphate, pH 7.5; (d) Tris phosphate, pH 8.0. Crystals of (b) have been tested by X-ray analysis. They diffracted up to 2.8 A ˚ but structure analysis could not been performed because the lifetime of the crystals during the measurements was extremely short even at temperatures between )140 and )160 °C. Fig. 4. Comparison of the secondary structure of the intact a-PEC and the peptide 2 obtained by nonspecific degradation. The alignment was performed with the structure viewer program WEBLAB VIEWER PRO and could be generated concerning the results of mass spectrometry and N-terminal amino acid sequencing (Table 1). The two N-terminal helices are responsible for the aggregation of a-PEC in solution. The mobile D pyrrole ring is marked by an arrow. 5050 G. Wiegand et al. (Eur. J. Biochem. 269) Ó FEBS 2002 The preparative separation of the peptides was achieved by the high performance ‘native’ PAGE, resolving five colored bands which have been analyzed by UV/visible spectroscopy, N-terminal sequencing and mass spectro- metry (Fig. 5, Table 1). The similarity of absorbance and fluorescence as well as the complementary phototransfor- mation of the peptides is indicative for the unchanged chromophore environment of the peptides. This observation could be confirmed by the comparative amino acid analyses and mass spectrometry. The chromopeptides 1–3 showed different N-terminal degradations, resulting in partially different charges of the peptides. Nevertheless, the elec- trophoretic separation cannot be explained solely by the peptide charges because bands with nearly the same charge (bands 1B, 1C and 2) migrated quite differently in the gel. It can be speculated that either structural factors or distinct aggregations of the peptides are responsible for the individ- ual migration behavior. The aggregation of the peptides 1A, 1B and 1C is shown from the behavior of these peptides during concentration by ultrafiltration. As shown in Table 1, they aggregate at pH 7.0, similar to the intact a-PEC subunit. Mass spectrometry of the PEC complexes and purified a-PEC was performed directly after isolation. The deter- mined molecular mass of the corresponding a-subunits agrees exactly with the calculated mass, including amino acids and the PVB chromophore. In contrast, within the crystals, two peptide masses differing by 16 Da have been detected. This fine but significant distinction reproducibly occurred in the crystal analyses and points to a modifi- cation of the chromoprotein during crystallization. Within the error limits, the difference of 16 Da corresponds well to an addition of oxygen. Although, the site of the oxidation could not be determined, it is probable that Cys98 and/or Cys99 of the a-subunit are partially oxygenated. The reaction mechanisms and conditions have not been investigated thoroughly, but it is an interesting result, especially regarding the photochemistry ofthetypesIandII[8,10]. Structure of peptide 2 The results summarized in the Table 1 clearly show that the two N-terminal helices are not necessary for the photo- chromism. Therefore, the molecular events accompanying the isomerization of the chromophore should be equivalent within the intact a-PEC and the derived peptides. The excellent solubility of peptides 2 and 3 at pH 7.0 recom- mended their employment for further studies such as crystallization and NMR spectroscopy. Unfortunately, depending on light, ionic strength and pH, the peptides are much more sensitive to degradation than the intact subunit. The reactions and their physical reasons have not been investigated systematically, however, all crystallization experiments failed and the peptides often lost their color Table 1. Comparison of the N-terminal sequences and molecular masses of the a-PEC peptides separated by ‘native’ polyacrylamide gel electrophoresis. Numbers in parentheses are minor components of the samples. Peptides Molecular mass N-terminus Molecular properties Photoactivity a-PEC 18 151.6 MKTPLTEAIAÆÆAADLRGSYLSÆÆNTELQAVFGRÆÆFNRARAGLEA Aggregating at pH 7.0 + Crystals of a-PEC 18 151.6 18 167.8 MKTPLTEAIAÆÆAADLRGSYLSÆÆNTELQAVFGRÆÆFNRARAGLEA Original molecule Modified molecule Peptide 1A 15 803.2 (15 473.8) ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆLQAVFGRÆÆFNRARAGLEA Aggregating at pH 7.0 + Peptide 1B 15 585.0 ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆAVFGRÆÆFNRARAGLEA Aggregating at pH 7.0 + Peptide 1C 15 157.0 ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆGRÆÆFNRARAGLEA Aggregating at pH 7.0 + Peptide 2 14 945.2 (14 797.2) ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ FNRARAGLEA ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆNRARAGLEA Soluble at pH 7.0 + Peptide 3 14 525.6 ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆÆ ÆARAGLEA Soluble at pH 7.0 + Fig. 5. High performance ‘native’ polyacrylamide electrophoresis of the a-PEC peptides. Cathode (–) is at the top and anode (+) at the bottom of the picture. All peptides show the ‘normal’ photoactivity suggesting a nearly unchanged chromophore environment. The peptides were analyzed by mass spectrometry and N-terminal amino acid sequen- cing. In both its E-andZ-configurations, peptide 2 was characterized further by NMR spectroscopy. Ó FEBS 2002 Analyses of a-phycoerythrocyanin peptides (Eur. J. Biochem. 269) 5051 (results not shown). An explanation for this behavior can be derived from the structural comparison of intact a-PEC and peptide 2 (Fig. 4). Within ‘monomeric’ PEC, hydrophobic interactions between the N-terminal helices of both the a-andb-subunits stabilize the complex [6]. At pH 7.0, similar interactions take place in the solutions of isolated a-PEC and the diffraction data suggest a ‘dimeric’ arrange- ment of the subunits within the crystals. Consequently, the association to ‘homodimers’ is proposed to be responsible for the enhanced stability of the intact a-subunit in contrast to that of the peptides. The low pH of 2.2 in 0.3% (v/v) formic acid, or at least the partial degradation of the two helices, prevents the interactions and reduces the aggrega- tion. However, complete loss of the helices or even more of the N-terminus significantly decreases the physical stability of the chromopeptides. Peptide 2 is characterized by a small stabilizing section of the second N-terminal helix and a sufficient solubility. Therefore, providing a good compro- mise between the two opposite molecular properties, this chromopeptide enabled light-dependent analysis by 1D NMR spectroscopy. Molecular alterations of the a-PEC peptide 2 demonstrated by NMR spectroscopy The purpose of the NMR study was not the detailed structural analysis of the two chromopeptide configura- tions. Moreover, the study should answer some important questions concerning the methodological knowledge and the molecular events depending on photochemistry: (a) Is peptide 2 suitable for further NMR studies? (b) Is the photochemistry of the chromophores accompanied by significant changes in the protein structure? (c) Is it possible to discern between chromophore and protein signals? (d) Is the photoconversion between the two states of the chromopeptide complete or incomplete, as indicated by the spectral data of the E-configuration (Fig. 2). Initial NMR spectroscopy was performed using the intact a-PEC, but protein aggregations caused extreme broadening of the signals. In contrast, the 1D NMR spectra of peptide 2 in its E-andZ-configuration, respectively, show the well separated peaks of a monomeric protein (Fig. 6). A reliable comparison between the spectra of one sample is possible as the light equipment enabled complementary irradiation within the NMR tube without changing the protein environment. For clarity, only the two important regions (NH and aliphatic) of the spectra are presented. The main differences between the spectral data of the E-andZ-configurations are emphasized by the E/Z-difference spectrum. Multiple spectral deviations in the height as well as the chemical shifts of the peaks can be seen. The various differences between nearly all regions of the spectra are indicative of parallel light-dependent structural changes of the peptide and the chromophore. The interpretation of the NMR spectra is rather difficult because protein and chromophore signals overlap. Obvi- ously, the presence of two peaks between 10 and 11 p.p.m. which do not change and their positions within the spectra suggests that they represent the two tryptophanes, Trp51 and Trp128, in the peptide [13,27], although an unusually shifted signal from another amino acid residue cannot be excluded. At least three peaks from the E-configuration and their slightly shifted negative counterparts from the Z-configuration are resolved in the aliphatic region of the difference spectrum. Because of the height and the sharpness of the peaks, they are assumed to be derived exclusively from the aliphatic residues of the distinct chromophore states. These signals probably reflect the isomerization and mobility of the D pyrrole ring. The dominant peaks of the peptide in the E-configuration show the enhanced mobility depending on reduced chromophore protein interactions in this state. The integration of well resolved peaks should enable an estimation of the state populations obtained by complementary illuminations. The protein peaks at )0.033 p.p.m and )0.099 p.p.m., as well as the protein peaks at 9.44 p.p.m and 9.38 p.p.m., can be attributed to the Z-andE-states, respectively. Integration of both pairs of peaks yields the ratios of state populations of Z/E ¼ 12%/88% for the E-state and approximately Z/E ¼ 65%/35% for the Z-state. These estimations are consistent within the various peaks of the NMR spectra but are contrary to the optical spectra, where only the E-form of a-PEC shows a significant amount of the complementary spectral state [9,10,21]. DISCUSSION This study presents the purification and molecular analyses of photoactive a-PEC peptides from phycobilisomes of M. laminosus. Preliminary results of crystallization and NMR spectroscopy offer reliable information on the relations between the protein backbone and the photo- chemically active chromophore of the peptides. Fig. 6. 1D NMR spectroscopy of a-PEC peptide 2 in 20 m M sodium-potassium phosphate, pH 7.2. The spectra were recorded after irradiation with light of 571 nm (E-configuration) and 503 nm (Z-configuration), respectively. To emphasize the spectral deviations the difference spectrum E-configuration–Z-configuration is presented. The spectra of the single sample have been recorded three times within a period of 3 months. Only the last spectrum, recorded after 3 months, showed significant deviations which could be attributed to a nonspe- cific degradation of the chromopeptide (results not shown). The chromophore peaks are marked by arrows and the integrated protein peaks are labeled by arrowheads. 5052 G. Wiegand et al. (Eur. J. Biochem. 269) Ó FEBS 2002 Methodological aspects In order to obtain high amounts of a-PEC, the purification methods have been scaled up without adversely changing the physical and chemical conditions of previous studies [18,19,21]. This means that the isolation media are almost identical, whereas the dissociation conditions for the purification of PEC complexes and a-PEC subunits, as well as the time consumption of all steps, have been optimized. In the ‘native’ PAGE of the isolated PEC fraction (Fig. 1), only the two naturally occurring PEC-linker complexes are present, confirming the brief dissociation and separation conditions [4]. The second important preparation step was that of hydrophobic interaction chromatography. Dissoci- ation of PEC and separation of the subunits take place within 2–3 h, which is extremely shortened in comparison with established separation methods [12,18,19]. a-PEC from M. laminosus was recently crystallized under white light, but the photoactive state of the proteins within the crystals has not been determined [19]. Therefore, it remains uncertain whether those crystals were composed of E-, Z-, or possibly both, states of the protein. However, the X-ray measurements of these crystals, as well as those of the crystals in this study, failed. Despite cryo-conditions, the molecular order of the crystals decreased rapidly during measurements. The reason for this is unknown, although the occurrence in both studies, as well as the markedly distinct crystallization behaviors of the E-andZ-states, point to the influence of light on the protein structure, even at low temperatures. It may be of interest that no cracks developed in the crystals during measurement. The considerable problem of the light sensitivity of a-PEC in all preparation, crystallization and measuring steps demands special light equipment. In crystallization, microscopic control and irradiation for NMR spectrometry, the light conditions have been optimized. Unequivocally, the X-ray measurements also need a protection light, and a long wavelength (650 nm) red light source is favored. Rapid degradation of a-PEC during all preparations has often been observed [19]. Certainly, one reason is the enhanced accessibility of isolated subunits to proteolytic enzymes. Nevertheless, other factors exist which are responsible for the degradation (see Results). The analyses of the peptides revealed various splitting positions of the amino acid chain. This variability cannot be explained by specific acid- or protease-catalyzed hydrolyses of the protein. Additionally, the stability of the chromopeptides decreases rapidly, depending on the presence and length of the two N-terminal helices, which has been proven by gel filtration after the last NMR measurements (3 months) of peptide 2 at pH 7.2. This sample showed a significant amount of degraded peptides (results not shown). With respect to all results, an ‘autolytical’ degradation at prefer- ential regions of the peptides can be suggested. Molecular features of a-PEC peptides The aggregation behavior and the tendency to degrade of isolated a-PEC strongly limited the investigation methods elucidating the molecular mechanisms of the photoconver- sion [10]. The isolation in formic acid enables working with high protein concentrations, although the influence of low pH between 2.0 and 2.5 on the molecular structure is not completely clear. Optical properties as well as photoactivity are almost equal in the pH range of 2.2–8.5 [9,10,12,19,21], so that a nearly unchanged protein structure around the chromophore must be assumed. Aggregation of a-PEC is assigned exclusively to the two N-terminal helices of the molecule that bind to each other via hydrophobic patches deviating from the association of the a-andb-subunits [6]. Subsequently, the dimers unspecifically associate to supra- molecular particles. Although, the excellent solubility of peptides 2 and 3 confirms this view, the explanation is not complete and the influence of low pH values also has to be considered. Low pH induces a partial and possibly a temporary unfolding of the N-terminal helices, depressing dimerization. A rapid degradation of these helices in formic acid which may be caused by their destabilization support this hypothesis. Thus, the physical stability of a-PEC is strongly correlated to the interactions of the N-terminal helices or at least parts of these helices (Table 1). The photochemistry of a-PEC peptides The photoactivity of the a-subunit strongly depends on its multiple protein interactions within the different association states of the PEC complexes [11,12,21]. The assembly of ‘monomeric’ and ‘trimeric’ complexes is accompanied by a decrease of the photochemistry from 100% of the isolated a-PEC to 8% for the ‘trimers’. Naturally, linker-free phycoerythrocyanin does not exist. Therefore, the slightly enhanced photoactivity of 11% of the linker-PEC com- plexes is of special interest. Structural and spectral results clearly show that some linker polypeptides are responsible for an increased flexibility of allophycocyanin and phyco- cyanin complexes [28; Reuter, unpublished results]. A similar behavior in the PEC-linker complexes would explain their relatively high photoactivity. Optical spectroscopy, as well as theoretical considera- tions, characterized the changes of the chromophore configuration on a substructural level [9,10,21,24,29–31]. Strong coupling of excited states within the chromophore and charge transfer states from the surrounding polar amino acid residues are assigned either to stabilize the E-and Z-configurations or to enable the fast photoinduced struc- tural changes [30]. The chromophore of the protein in the E-configuration also shows pronounced Z-characteristics spectrally (Fig. 2), suggesting either a higher mobility or the existence of different intermediate states of the D pyrrole ring [24,31]. The role of the apoprotein conformation in the spectral behavior of the chromophore is unknown because almost all applications focus on the chromophore and its neighboring amino acids. A first indication for considerable structural deviations of a-PEC in the E-andZ-states can be derived from their crystallization behavior. The E-state crystallizes under various conditions whereas crystals, or at least microcrys- tallization, of the Z-state have never been observed. This result correlates well with the NMR data, where the protein peaks of the molecule in the E-conformation are much more homogenous than that of the Z-conformation. On the other hand, the mobility of some aliphatic groups of the E-chromophore are clearly increased compared with those of the Z-chromophore (Fig. 6). The NMR data can be interpreted as stabilization of the Z-chromophore configur- ation by an enhanced protein flexibility. This situation has Ó FEBS 2002 Analyses of a-phycoerythrocyanin peptides (Eur. J. Biochem. 269) 5053 actually been simulated by molecular dynamics and was described as oscillation of the chromophore and its environment [30]. In contrast, the protein in the E-state is more rigid, although the D pyrrole ring of the chromophore moves between its E and the Z positions. At present, the function of the photochemistry in PEC is uncertain because the analysis in the environment of the phycobilisomes is not currently possible. According to evolution studies on the phycobiliproteins of cyanobacteria and rhodophyceae, PEC is the youngest member of this protein family [32]. Unequivocally, a-PEC is not a sponta- neous mutation of a phycocyanin gene because two special lyases are involved in the synthesis and attachment of the chromophore [33,34]. Concerning the light harvesting, the advantage of PEC complexes compared with phycocyanin complexes is the broadening of the phycobilisome absorb- ance in the green light gap, whereas the photochemistry of a-PEC may function in a radiationless energy dissipation. However, the missing fluorescence of PEC in intact phycobilisomes and different adapted cells of M. laminosus support this suggestion [4]. ACKNOWLEDGEMENTS This work was financially supported by the Deutsche Forschungsg- emeinschaft, Sonderforschungsbereich 533 (projects A1, A2, A3). The authors wish to thank K H. Mann for N-terminal amino acid analyses andF.SiedlerandS.Ko ¨ rner for mass spectrometry. REFERENCES 1. Glazer, A.N. (1989) Light guides – directional energy transfer in a photosynthetic antenna. J. Biol. Chem. 264, 1–4. 2. Sidler, W.A. (1994) Phycobilisome and Phycobiliprotein Struc- tures. In The Molecular Biology of Cyanobacteria (Bryant, D.A., ed.), pp. 139–216, Kluwer Academic Publishers, Dortrecht, the Netherlands. 3. Bryant, D.A. (1982) Phycoerythrocyanin and phycoerythrIn Properties and occurrence in cyanobacteria. J. Gen. Microbiol. 128, 835–844. 4. Reuter, W. & Nickel-Reuter, C. (1993) Molecular assembly of the phycobilisome from the cyanobacterium Mastigocladus laminosus. J. Photochem. Photobiol., B 18, 51–66. 5. Reuter, W. & Mu ¨ ller, C. (1993) Adaptation of the photosynthetic apparatus of cyanobacteria to light and CO 2 . J. Photochem. Photobiol., B 21, 3–27. 6. Du ¨ rring, M., Huber, R., Bode, W., Ru ¨ mbeli, R. & Zuber, H. (1991) Refined three-dimensional structure of phycoerythrocyanin from the cyanobacterium Mastigocladus laminosus at 2.7 A ˚ . J. Mol. Biol. 211, 633–644. 7. Bjo ¨ rn, L.O. (1979) Photoreversibly photochromic pigments in organism: properties and role in physiological light percecption. Quart. Rev. Biophys. 12, 95–113. 8. Hong, Q., Zhao, K H. & Scheer, H. (1993) Two different types of photochemistry in phycoerythrocyanin a-subunit. Photochem. Photobiol. 58, 745–747. 9. Zhao, K H., Haessner, R., Cmiel, E. & Scheer, H. (1995) Type I reversible photochemistry of phycoerythrocyanin involves Z/E-isomerization in a-84 phycoviolobilin chromophore. Biochim. Biophys. Acta 1228, 235–243. 10. Zhao, K H. & Scheer, H. (1995) Type I and type II reversible photochemistry of phycoerythrocyanin a-subunit from Mastigo- cladus laminosus both involve Z, E isomerization of phycoviolo- bilin chromophore and are controlled by sulfhydryls in apoprotein. Biochim. Biophys. Acta 1228, 244–253. 11. Siebzehnru ¨ bl, S., Fischer, R., Kufer, W. & Scheer, H. (1989) Photochemistry of phycobiliproteins: reciprocity of reversible photochemistry and aggregation in phycoerythrocyanin from Mastigocladus laminosus. Photochem. Photobiol. 49, 753– 761. 12. Parbel, A. (1997) Charakterisierung von Phycobiliprotein-Link- erkomplexen aus dem Phycobilisom von. Mastigocladus laminosus. Vergleich nativer und u ¨ berexprimierter Linker.PhDThesis, Ludwig-Maximilians-Universita ¨ t, Mu ¨ nchen, Germany. 13. Fu ¨ glistaller, P., Suter, F. & Zuber, H. (1983) The complete amino- acid sequence of both subunits of phycoerythrocyanin from the thermophilic cyanobacterium Mastigocladus laminosus. Hoppe- Seyler’s Z. Physiol. Chem. 364, 691–712. 14. Bishop, J.E., Rapoport, H., Klotz, V., Chan, C.F., Glazer, A.N., Fu ¨ glistaller, P. & Zuber, H. (1987) Chromopeptides from phy- coerythrocyanin. Structure and linkage of the three bilin groups. J. Am. Chem. Soc. 109, 875–881. 15. Hughes, J. & Lamparter, T. (1999) Procaryotes and phytochrome. The connection to chromophores and signaling. Plant Physiol. 121, 1059–1068. 16. Neff, M.M., Fankhauser, C. & Chory, J. (2000) Light: an indicator of time and place. Genes Dev. 14, 257–271. 17. Reuter, W. & Wehrmeyer, W. (1988) Core substructure in Mas- tigocladus laminosus phycobilisomes. I. Microheterogeneity in two of three allophycocyanin core complexes. Arch. Microbiol. 150, 534–540. 18. Fu ¨ glistaller, P., Widmer, H., Sidler, W., Frank, G. & Zuber, H. (1981) Isolation and characterization of phycoerythrocyanin and chromatic adaptation of the cyanobacterium Mastigocladus laminosus. Arch. Microbiol. 129, 268–274. 19. Zhang, Z Y., Zhou, M., Zhao, K H., Zhang, J P., Chang, W R. & Liang, D K. (2000) Purification and crystallization of phy- coerythrocyanin a-subunit from Mastigocladus laminosus. Acta Biophysica Sinica 16, 667–672. 20. Kume, N., Isono, T. & Katoh, T. (1982) Stability of cyano- bacterial phycobilisomes in reference to their concentration. Photobiochem. Photobiophys. 4, 25–37. 21. Parbel, A., Zhao, K H., Breton, J. & Scheer, H. (1997) Chro- mophore assignment in phycoerythrocyanin from Mastigocladus laminosus. Photosynth. Res. 54, 25–34. 22. Schnackenberg, J., Than, M.E., Mann, K H., Wiegand, G., Huber, R. & Reuter, W. (1999) Amino acid sequence, crystal- lization and structure determination of reduced and oxidized cytochrome c 6 from the green alga Scenedesmus obliquus. J. Mol. Biol. 290, 1019–1030. 23. Plateau, P. & Gueron, M. (1982) Exchangeable proton NMR without base-line distortion, using new strong-pulse sequences. J. Am. Chem. Soc. 104, 7311–7317. 24. Kneip, C., Parbel, A., Foerstendorf, H., Scheer, H., Siebert, F. & Hildebrandt, P. (1998) Fourier transform near-infrared resonance raman spectroscopic study of the a-subunit of phycoerythrocyanin and phycocyanin from the cyanobacterium Mastigocladus lami- nosus. J. Raman Spectrosc. 29, 939–944. 25. Foerstendorf, H., Benda, C., Ga ¨ rtner, W., Storf, M., Scheer, H. & Siebert, F. (2001) FTIR Studies of phytochrome photoreactions reveal the C–O bands of the chromophore: consequences for its protonation states, conformation and protein interaction. Biochemistry 40, 14952–14959. 26. Jauregui-Adell, J. & Marti, J. (1975) Acidic cleavage of the aspartyl-proline bond and the limitations of the reaction. Anal. Biochem. 69, 468–473. 27. Eberlein, M. & Kufer, W. (1990) Genes encoding both subunits of phycoerythrocyanin, a light-harvesting biliprotein from the cya- nobacterium Mastigocladus laminosus. Gene 94, 133–136. 28. Reuter, W., Wiegand, G., Huber, R. & Than, M.E. (1999) Structural analysis at 2.2 A ˚ of orthorhombic crystals present the asymmetry of the allophycocyanin-linker complex, APÆL 7:8 C , from 5054 G. Wiegand et al. (Eur. J. Biochem. 269) Ó FEBS 2002 phycobilisomes of Mastigocladus laminosus. Proc. Natl Acad. Sci. USA 96, 1363–1368. 29. Hucke, M., Schweitzer, G., Holzwarth, A.R., Sidler, W. & Zuber, H. (1993) Studies of chromophore coupling in isolated phycobi- liproteins. IV. Femtosecond transient absorption study of ultrafast excited state dynamics in trimeric phycoerythrocyanin complexes. Photochem. Photobiol. 57, 76–80. 30. Scharnagel, C. & Fischer, S.F. (1993) Reversible photochemistry in the a-subunit of phycoerythrocyan: characterization of chro- mophore and protein by molecular dynamics and quantum chemical calculations. Photochem. Photobiol. 57, 63–70. 31. Foerstendorf, H., Parbel, A., Scheer, H. & Siebert, F. (1997) Z, E isomerization of the a-84 phycoviolobilin chromophore of phy- coerythrocyanin from Mastigocladus laminosus investigated by fourier-transform infrared difference spectroscopy. FEBS Lett. 402, 173–176. 32. Apt, K.E., Collier, J.L. & Grossman, A.R. (1995) Evolution of the phycobiliproteins. J. Mol. Biol. 248, 79–96. 33. Zhao, K H., Deng, M G., Zheng, M., Zhou, M., Parbel, A., Storf, M., Meyer, M., Strohmann, B. & Scheer, H. (2000) Novel activity of a phycobiliprotein lyase: both the attachment of phy- cocyanobilin and the isomerization to phycoviolobilin are cata- lyzed by the proteins PecE and PecF encoded by the phycoerythrocyanin operon. FEBS Lett. 469, 9–13. 34. Storf, M., Parbel, A., Meyer, M., Strohmann, B., Scheer, H., Deng, M G., Zheng, M., Zhou, M. & Zhao, K H. (2001) Chromphore attachment to biliproteins: specifity of PecE/PecF, a lyase-isomeraseforthephotoactive3 1 -Cys-a84-phycoviolobilin chromophore of phycoerythrocyanin. Biochemistry 40, 12444– 12456. Ó FEBS 2002 Analyses of a-phycoerythrocyanin peptides (Eur. J. Biochem. 269) 5055 . Purification, crystallization, NMR spectroscopy and biochemical analyses of a-phycoerythrocyanin peptides Georg Wiegand 1 , Axel Parbel 2 *,. In both its E-andZ-configurations, peptide 2 was characterized further by NMR spectroscopy. Ó FEBS 2002 Analyses of a-phycoerythrocyanin peptides (Eur.