Characterization of phycoviolobilin phycoerythrocyanin-a84-cystein- lyase-(isomerizing) from Mastigocladus laminosus Kai-Hong Zhao 1 , Dong Wu 1 , Lu Wang 1 , Ming Zhou 1 , Max Storf 2 , Claudia Bubenzer 2 , Brigitte Strohmann 2 and Hugo Scheer 2 1 College of Life Science and Technology Huazhong University of Science and Technology, Wuhan, Hubei, China, 2 Botanisches Institut, Universita ¨ tMu ¨ nchen, Germany Cofactor requirements and enzyme kinetics have been studied of the novel, dual-action enzyme, the isomerizing phycoviolobilin phycoerythrocyanin-a84-cystein-lyase(PVB- PEC-lyase) from Mastigocladus laminosus, which catalyses both the covalent attachment of phycocyanobilin to PecA, the apo-a-subunit of phycoerythrocyanin, and its isomeri- zation to phycoviolobilin. Thiols and the divalent metals, Mg 2+ or Mn 2+ , were required, and the reaction was aided by the detergent, Triton X-100. Phosphate buffer inhibits precipitation of the proteins present in the reconstitution mixture, but at the same time binds the required metal. Kinetic constants were obtained for both substrates, the chromophore (K m ¼ 12–16 l M , depending on [PecA], k cat  1.2 · 10 )4 Æs )1 ) and the apoprotein (K m ¼ 2.4 l M at 14 l M PCB, k cat ¼ 0.8 · 10 )4 Æs )1 ). The kinetic analysis in- dicated that the reconstitution reaction proceeds by a sequential mechanism. By a combination of untagged and His-tagged subunits, evidence was obtained for a complex formation between PecE and PecF (subunits of PVB-PEC- lyase), and by experiments with single subunits for the pre- valent function of PecE in binding and PecF in isomerizing the chromophore. Keywords: chromophore; cyanobacteria; enzymology; pho- tosynthesis; phycobilin isomerization; phycobilin lyase; phycobiliprotein synthase; thiol addition. Phycobilisomes are the major photosynthetic antenna complexes of cyanobacteria and red algae [1,2]. They harvest light in the green-gap of chlorophyll absorption, and transfer excitation energy with high quantum efficiency to the photosynthetic reaction centers, mainly photosystem (PS)II. Phycobilisomes are composed of phycobiliproteins, which absorb light, and linker proteins, which organize the former into the phycobilisome and modulate their absorp- tions. Some of the linkers also carry bilin chromophores. Cyanobacterial and red-algal biliproteins are generally trimers of an a/b-heterodimer. a-andb-subunits are closely related proteins, carrying 1–4 covalently bound chromoph- ores, the phycobilins. Based on their absorption spectra properties, phycobiliproteins have originally been classified into three major groups: allophycocyanins, phycocyanins (PC), and phycoerythrins (PE). The former two carry mainly phycocyanobilin (PCB) chromophores with a single covalent bond linking C-3 1 to cysteine residues of the apoproteins, while PE is characterized by phycoerythrobilin (PEB) chromophores. However, the type of chromophore as well as the mode of binding can be considerably more complex [3,4]. Urobilin chromophores are frequently found in PE and PC from marine cyanobacteria. Phycoerythro- cyanin (PEC) carries a phycoviolobilin chromophore. PEC is a light-harvesting component of the phycobilisome in some filamentous, N 2 -fixing cyanobacteria. However, un- like the other biliproteins, PEC shows a photochemistry reminiscent of the sensory photoreceptors, phytochromes (see for example [5–9]), which has been attributed to the phycoviolobilin (PVB) chromophore [10,11]. Of the four cyanobacterial and red-algal phycobilins, PCB and PEB possess a D3,3 1 -ethylidene group. They are synthesized from haem by ring opening at C-5 of the tetrapyrrole and several reduction steps, and then attached to the apoproteins by addition of a cystein thiol to the ethylidene group [3,4,12–14]. PCB and PEB can add thiols spontaneously and reversibly, including cysteines of Correspondence: K H. Zhao, College of Life Science and Technology, Huazhong University of Science and Technology, Wuhan 430074, Hubei, P.R. China. Fax: +86 27 8754 1634, Tel.: +86 27 8754 1634, E-mail: khzhao@163.com Hugo Scheer, Botanisches Institut, Universita ¨ tMu ¨ nchen, Menzinger Str. 67, D-80638 Mu ¨ nchen, Germany. Fax: +49 89 17861 185, Tel.: +49 89 17861 295, E-mail: scheer-h@botanik.biologie.uni-muenchen.de Abbreviations: DDA xxx/yyy , amplitude of photochemical signal with difference maxima at xxx and yyy nm, normalized to maximum absorption (see ref [11]. for details); DME, dimethylester; PCB, phycocyanobilin; PEB, phycoerythrobilin; PEC, phycoerythrocyanin; PVB-PEC-lyase, phycoviolobilin phycoerythrocyanin-a84-cystein- lyase (isomerizing); PecA, apoprotein of a-PEC; PecE, PecF, subunits of PVB-PEC-lyase; PUB, phycourobilin; PVB, phycoviolobilin; (There are two terms for the chromophore in the literature: phycobi- liviolin [41] and phycoviolobilin [11]; the latter is used because it is analogous to the names of the major phycobilins, viz. phycocyano- and phycoerythrobilin.) PC, C-phycocyanin; PEC, phycoerythro- cyanin; a-PEC, chromophorylated a-PEC; TX-100, Triton X-100. Enzymes: phycoviolobilin phycoerythrocyanin-a84-cystein-lyase (PVB-PEC-lyase; isomerizing). This name has been submitted to ENZMES, as an enzyme of the subclass 4.4.1, as an alternative name we proposed holo-a-phycoerythrocyanin synthase, in analogy to the cytochrome synthase 4.4.1.17. Note: the names of all chromophores refer to the free chromophores, while the chromophores attached to the apoproteins are characterized as addition products. (Received 21 April 2002, revised 23 July 2002, accepted 26 July 2002) Eur. J. Biochem. 269, 4542–4550 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03148.x apo-biliproteins, forming a relatively stable thioether bond [3,5,15–17]. However, the regio- and stereo-specifically correct attachment has been demonstrated only for the phytochromes (see [14]) and for a single site in PC, b-84 [18,19]. Attachment to the cys-a-84 of CpcA, the a-PC apoprotein, is catalysed enzymatically by heterodimeric lyases [20,21]. Genes encoding homologous proteins are known from many cyanobacteria, but the functions of most are unknown. Biliproteins are also known to have up to five binding sites per monomer, and little is currently known about the attachment to sites other than cys-a-84 (reviewed in [22]). Little information is available about the biosynthesis of the other two chromophores, PVB and phycourobilin (PUB). These free chromophores have not yet been isolated from any source, they are only known as protein-bound 3 1 - thiol adducts. Due to the presence of a D2,3-double bond, which precludes a second one at D3,3 1 , their mode of attachment to the apoproteins also must be different from that of PCB and PEB. For PVB, the problem has recently been clarified by the identification of a new enzymatic activity of the two lyase subunits (PecE, PecF), whose genes are located on the pec operon. In addition to catalysing the covalent attachment of PCB to cys-a-84 of PecA, the PEC a-subunit, they promote a concomitant isomerization [23,24]. The result of this intriguing double action is a-PEC, with the correct 3 1 -cys-PVB chromophore attached to PecA. A similar reaction sequence would lead from PEB to the 3 1 -cys-PUB chromophore present in many PE. However, no such enzyme is currently known, and the PVB- PEC-lyase (PecE/F) (previously termed lyase-isomerase) does not accept the PEB as substrate. Intrigued by its unusual photochemistry, we became interested in protein engineering a-PEC. One goal is to establish the structural basis of the highly reversible photochemistry of a-PEC and the protein dynamics related to the transformation, the other is to evaluate the potential of the relatively small chromoprotein as a photo switch. The isomerizing PVB-PEC-lyase is crucial to this project: it allows us to modify separately both the prosthetic group, i.e. the chromophore, and the apoprotein of a-PEC in a directed manner, and then to reconstitute the chromoprotein in vitro. The a-PEC syntase consists of two proteins, PecE and PecF, whose genes are encoded in the pec-operon down- stream from the structural (pecB, A) and linker genes (pecC). Our previous experiments showed that under catalysis of the crude extract of heterologously (Escherichia coli) over-expressed PecE and PecF, PCB can be converted to PVB, and bound covalently to apo-a-PEC to give native a-PEC. We now report on the preparation of the subunits of PVB-PEC-lyase possessing His 6 -tags at the N terminus (to facilitate purification) and on their enzymatic char- acterization. MATERIALS AND METHODS Overexpression of His 6 -PecA, His 6 -PecE, and His 6 -PecF The genes pecA, pecE,andpecF were cloned from Mastigocladus laminosus (Fischerella spec.) with vector pBluescript (Stratagene), yielding plasmids pBlu-pecA, pBlu-pecE, and pBlu-pecF, respectively. All constructions were verified by sequencing. These genes were subcloned into vector pET30a (Novagen) using the EcoRV and HindIII restriction sites (pecA)orEcoRV and XhoI restriction sites (pecE, pecF). pBlu-pecA, pBlu-PecE and pBlu-PecF were cleaved with Smal IandXhoI, and the released genes were ligated to the large pET30a fragment. Purification of His 6 -PecA, His 6 -PecE, and His 6 -PecF His-tagged PecA, E and F were purified separately by metal ion chelating affinity chromatography on chelating seph- arose (fast flow; Amersham Pharmacia Biotech AB, according to the supplier’s protocol) charged with Ni 2+ . The E. coli [strain BL21(DE3)] cells containing recombinant pET30a were grown in Luria–Bertani medium at 37 °C, and harvested 5 h after induction with isopropyl thio-b- D - galactoside (IPTG). The cells (usually from 1 L of culture) were washed twice with distilled water, and then suspended in 30 mL start buffer (20 m M potassium phosphate buffer pH 7.2 containing 0.5 M NaCl). The suspension was sonicated (Branson model 450 W, 30 min, 45 W) to break the cells, and then centrifuged for 30 min at 12000 g.The supernatant was loaded directly onto the Ni 2+ chelating affinity column. After washing with 5 column vols of start buffer to remove untagged proteins, His 6 -PecA, His 6 -PecE, or His 6 -PecF were eluted with stripping buffer (20 m M potassium phosphate buffer, pH 7.2 containing 100 m M EDTA, 0.5 M NaCl). The eluent was dialysed three times against 50 m M potassium phosphate buffer pH 7.2, con- taining 0.5 M NaCl, to remove Ni 2+ and EDTA. Optimi- zation experiments showed that the latter buffer prevents the otherwise ready precipitation of the His-tagged proteins. An alternative protocol resulted in proteins which can be stored better and for longer: the His-tagged proteins were first dialysed against 50 m M potassium phosphate buffer containing 0.5 M NaCl, pH 7.2, and then twice against the same buffer containing also 1 m M 2-mercaptoethanol. Finally, His 6 -PecA was stored at )20 °C; His 6 -PecE, and His 6 -PecF were mixed with an equal volume of glycerol before storing at )20 °C. These proteins did not show any loss of activity after storage for 1 year at )20 °C. A 1-L culture of E. coli yielded routinely  100 mg of PecA, 50 mg of PecE or 30 mg of PecF. PCB preparation PCB was prepared as described before [23]. Typical reconstitution of PCB with His 6 -PecA under catalysis of His 6 -PecE, and His 6 -PecF The reconstitution system consists of PCB, His 6 -PecA, His 6 - PecE, His 6 -PecF, and 2-mercaptoethanol. The enzyme reaction was carried out at 37 °C for times between 15 min and 3 h, or at room temperature for 1–12 h. PCB was added in dimethylsulfoxide solution, to a final concen- tration of 1% (v/v) dimethylsulfoxide. His 6 -PecA, His 6 - PecE, and His 6 -PecF were added to final at concentrations of 17–86, 6.6–33, and 8.9–44 l M , respectively. The final concentration of the other components were: potassium phosphate buffer, 15–20 m M ; NaCl, 150–200 m M ; 2-merca- ptoethanol, 5 m M ; glycerol, 10% (v/v). In addition, 2-mercaptoethanol and a divalent metal (Mn 2+ or Mg 2 ) are necessary for the activity of His 6 -PecE and His 6 -PecF. Ó FEBS 2002 Enzymology of isomerizing phycoviolobilin lyase (Eur. J. Biochem. 269) 4543 The optimum concentration of 2-mercaptoethanol is 5 m M . The activating metals (Mg 2+ or Mn 2+ )hadoptimumcon- centrations of 5 and 3 m M , respectively. Mn 2+ is favoured, as it stabilized the PVB-PEC-lyase at 37 °C. The pH of the reconstitution system was adjusted to 7.5 by addition of Tris/HCl (1 M , pH 7.5) to an end concentration of 100 m M (see Results). Finally, the following detergents at end con- centrations of 1% (v/v) were shown to improve the reaction: Triton X-100 (usually used), Nonidet P-40, or Tween-20. In this work the His 6 -PecA, His 6 -PecE, and His 6 -PecF were used at 15–25 l M , and the chromophore substrate, PCB, was used at 25 l M , unless stated otherwise. At the low concentrations used for the experiments, which allowed easy monitoring by spectrophotometry, PCB is otherwise liable to oxidation by air and precipitation, resulting in loss of the chromophore and formation of by-products. Spectroscopy TheenzymereactionwasmonitoredwithaUV-VIS spectrophotometer (Perkin-Elmer model Lamda2). The product absorption (Z-a-PEC) was monitored at 570 nm. For better characterization, the reversible photoreaction of the reconstituted product was routinely quantitated by its DDA as detailed in [11]. Reconstituted His 6 -a-PEC has a DDA of 100–110% (see Results), the measured DDA therefore almost equals the absorption of the correctly reconstituted product, a-PEC (127%, [10,24]). The extinc- tion coefficient of His 6 -a-PEC was taken as that of the untagged protein (e 562 ¼ 1.0 · 10 5 Æ M )1 Æcm )1 [11]). Intermediates of the enzyme reaction Reconstitution reactions were carried out as above, but stopped after 1 h at 37 °C by the addition of 0.2% trifluroacetic acid (v/v). After further addition of 2-propanol (70%, v/v), the mixture was centrifuged to remove any precipitates. The supernatants were injected into HPLC (RP 18 column) and analysed in stream with the diode-array detector (J & M model Tidas). PCB and protein concentration determinations PCB concentration was determined spectroscopically using an extinction coefficient e 690 ¼ 37.9 m M )1 Æcm )1 in meth- anol/2% HCl. The protein concentration was assayed with protein assay reagent (Bio-Rad) according to the instruc- tions given by the supplier using BSA as a standard. SDS/PAGE was performed according to Laemmli [25]. RESULTS AND DISCUSSION Overexpression of His 6 -PecA, His 6 -PecE, and His 6 -PecF All three genes could be overexpressed effectively in the vector pET30a. The over-expressed His-tagged proteins required sonification of the cell suspension for relatively long times (30 min to bring  90% into solution). After centrifugation, the supernatant containing His 6 -PecA can already be used for many reconstitution experiments [24], this solution is also used for purification via Ni 2+ chelating chromatography. His 6 -PecE and His 6 -PecF are more soluble. After sonification of the respective E. coli cells, they resided quantitatively in the supernatant, which can be used for reconstitution [24] or subsequent purifi- cation. Purification and conservation of His 6 -PecA, His 6 -PecE, and His 6 -PecF In previous experiments, the subunits of PVB-PEC-lyase, PecE and PecF were over-expressed using the pGEMEX vector [24]. Attempts to purify crude extracts of the lyase (ammonium sulfate precipitates) proved difficult and resul- ted in loss of activity. Therefore, no further attemp was made to improve the purification methods; instead we concentrated on over-expression by switching to the pET30a vector. The resulting His 6 -PecA, His 6 -PecE, and His 6 -PecF were easily purified by Ni 2+ chelating affinity chromatography [23]. A 1-L culture yielded routinely  100 mg PecA, 50 mg PecE or 30 mg PecF, which by affinity chromatography were concentrated to 10 mgÆmL )1 , without loss of activity. However, the proteins are liable to precipitation. Solu- bility was greatly enhanced by using potassium phosphate buffer containing high concentrations of NaCl (0.5 M ). Phosphate is critical because it interacts with the cofactors, Mg 2+ and Mn 2+ (see below), but this effect could be largely compensated by increasing the concentration of these ions. It is convenient to strip the His 6 -tagged proteins with potassium phosphate buffer (20 m M , pH 7.2) containing EDTA (100 m M ), and NaCl (0.5 M ), then the eluent can be fractionated according to the colour of the eluent due to Ni 2+ . In this case, exhaustive dialysis is necessary to remove any Ni 2+ , which quenches the enzyme (see below). The over-expressed His-tagged PecA, PecE, PecF are stable, if kept frozen at )20 °C. There was no loss of activity over 1 year. Repetitive freezing and thawing should be avoided, as it causes the purified proteins to precipitate, particularly His 6 -PecE and His 6 -PecF. Very stable preparations were obtained by adding an equal volume of glycerol before freezing. The purified His 6 -PecA can be also conserved by this method, but in this case care has to be taken keep the glycerol concentration < 10% (v/v), as more decreased the PVB-PEC-lyase activity. Optimal conditions for a-PEC reconstitution The enzyme reaction was shown previously to require several cofactors [23]. These requirements were now tested in more detail. Metal specificity. Activation of the PVB-PEC-lyase is more effective with Mn 2+ than with Mg 2+ (Table 1), although without Mn 2+ andeveninthepresenceof 0–50 m M EDTA, the PVB-PEC-lyase had still 21% activity. The optimum concentration of Mn 2+ is 3 m M ,thatof Mg 2+ is 5 m M . Higher concentrations of Mn 2+ are critical. Firstly, when Mn 2+ wasusedasanactivator,the concentration of 2-mercaptoethanol needed to be a little higher than that of Mn 2+ , otherwise the Mn 2+ was less effective (see below). Secondly, Mn 2+ concentration of 25 m M completely inhibited the PVB-PEC-lyase. By com- parison, the same concentration of Mg 2+ (25 m M ) still resulted in 62% of the maximum activity. Thirdly, the metal ions differ in the effect of EDTA on the PVB-PEC-lyase. 4544 K H. Zhao et al. (Eur. J. Biochem. 269) Ó FEBS 2002 It inhibited catalysis by Mn 2+ much more rapidly than by Mg 2+ . Last but not least, when Mn 2+ wasusedasan activator, the Mn 2+ /EDTA complex accelerated the oxi- dation of chromophore, resulting in rapid loss of chromo- phore and complex reconstitution mixtures. While this can be compensated in analytical assays by an excess of the lyase, it is problematic for preparative reconstitutions. We also found out that it was beneficial to prepare concentrated stock solutions of Mn 2+ ,tostorethemat)20 °C, and to add the correct amount of Mn 2+ immediately before starting the enzyme reaction. Possibly, this prevents oxida- tion and dimerization of the metal ion, but this hypothesis was not pursued in detail. Several other divalent metals were tested for catalysis (Table 1). Besides Mn 2+ and Mg 2+ , only Ca 2+ activated the isomerase activity of PVB-PEC-lyase, but was less effective than the former. All other metal ions tested (Fe 2+ , Co 2+ ,Ni 2+ ,Cu 2+ ,Zn 2+ )resultedininactivationof enzymatic isomerization. Because the enzyme is moderately active in the absence of the activating metals, Mg 2+ or Mn 2+ (Table 1), but the activity is completely lost in the presence of the other metals tested, this indicates a genuine inactivation by the latter. Because Ni 2+ was used to bind His 6 -tagged proteins during metal chelating chromatogra- phy, any Ni 2+ eluted in the process was removed from isolated His 6 -tagged PecA, PecE, and PecF, by exhaustive dialysis against the potassium phosphate/NaCl reconstitu- tion buffer. With some metals, unwanted side reactions were observed in addition: Fe 2+ ,Cu 2+ ,andZn 2+ accelerated chromophore oxidation; Co 2+ formed a complex with 2-mercaptoethanol absorbing around 470 nm; and in the presence of Ni 2+ , a broad, unstructured absorption formed in the 610–650 nm region. Both Mn 2+ and Mg 2+ are ubiquitous chelators of nucleotides and cofactors of many related enzymes. The catalysis by Mn 2+ was therefore tested in the presence of ATP and or GTP (data not shown). However, neither increased the activity of catalysis by His 6 -PecE/His 6 -PecF. Because metals can complex linear tetrapyrroles, the effect of the metals was investigated on the absorption spectrum of PCB: addition of Mn 2+ caused no change to the visible absorption spectrum, irrespective of the presence or absence of the His 6 -PecE and His 6 -PecF (data not shown). As the chromophore spectrum is very sensitive to environmental changes (see for example the effect of Triton X-100 discussed below), these results suggest that Mn 2+ acted on the PVB-PEC-lyase, and not (or only transitorily) on the conformation of PCB. Thiols 2-mercaptoethanol or thiols like such as dithiothreitol are required for the isomerization reaction of the lyase: without, only the PCB addition product was formed, but no PVB- His 6 -a-PEC (Table 1). Also the spontaneous addition (no enzymes added) of PCB yielding the cys-a84-PCB-adduct, proceeds in the absence of 2-mercaptoethanol or other thiols. However, too much 2-mercaptoethanol will cause the loss of chromophore, in a reaction requiring oxygen. When Mg 2+ was used as the activator, the optimal concentration of 2-mercaptoethanol is 5 m M ,withMn 2+ it is 3 m M .The effect of thiols is specific, they could not be replaced by other biological reductants such as NADPH or ascorbic acid (data not shown). Other factors influencing activity NaCl is beneficial to the reconstitution by preventing the precipitation of the over-expressed proteins. However, it proved inhibitory at high concentrations. The activity of the His 6 -PecE and His 6 -PecF was not noticeably affected up to 250 m M NaCl, but decreased to 50% in 500 m M . A similar optimum was found with potassium phosphate, which is needed to dissolve the purified His 6 -tagged proteins; but this buffer decreases the effect of the activators, Mn 2+ or Mg 2+ , by 20% as compared with Tris/HCl buffer. This is most probably due to formation of metal complexes. To balance these effects, it proved best practice to use a mixed buffer system consisting of one volume of potassium phosphate (50 m M ) containing NaCl (0.5 M ), and two volumes of Tris/ HCl (150 m M ), resulting in final concentrations of 17, 170 and 100 m M , respectively. Under these conditions, the lyase has an optimal pH at around 7.5–7.8 (Fig. 1). As the bilins become more liable to oxidation at higher pH [26], a pH £ 7.5 was favoured, and usually buffers of pH 7.5 were used in this work. Temperature. The lyase requires relatively high tempera- tures (Table 2). With Mg 2+ , the reaction time at room temperature of 1 h, can be reduced to 10 min at 37 °C. However, room temperature is recommended in the absence of TX-100, because the proteins tend to precipitate at 37 °C. Table 1. Relative activities of the lyase, His 6 -PecE and His 6 -PecF, depending on the presence and concentrations of 2-mercaptoethanol and divalent metals. Relative activities were determined by the type I photochemical activity of the product according to Zhao et al. [11]. Added cofactors (concentrations [mM]) Relative activity (%) ME (0), or ME (0) and Mn 2+ (3) 0 ME (5) 21 ME (5), EDTA (5–50) 21 ME (5), EDTA (5), Mg 2+ (5) 70 ME (5), EDTA (10), Mg 2+ (5) 21 ME (5), EDTA (5–10), Mn 2+ (3) 2 a ME (2.5), Mn 2+ (1.2) 55 ME (5), Mn 2+ (5) 64 ME (5), Mn 2+ (3) 100 ME (10), Mn 2+ (5) 87 ME (25), Mn 2+ (5) 71 ME (50), Mn 2+ (5) 42 ME (5), Mn 2+ (10) 13 ME (5), Mn 2+ (25) 0 ME (5), Mg 2+ (2.5) 58 ME (5), Mg 2+ (5) 80 ME (5), Mg 2+ (10) 74 ME (5), Mg 2+ (25) 51 ME (5), Mg 2+ (5) + Ca 2+ (5) 73 ME (5), Ca 2+ (5) 68 ME (5), Fe 2+ (5) 0 ME (5), Cu 2+ (5) 0 ME (5), Zn 2+ (5) 0 ME (5), Co 2+ (5) 0 ME (5), Ni 2+ (5) 0 a Acceleration of chromophore oxidation by Mn 2+ –EDTA. Ó FEBS 2002 Enzymology of isomerizing phycoviolobilin lyase (Eur. J. Biochem. 269) 4545 In the presence of TX-100 (1% v/v), the temperature can be increased without precipitation to 37 °C(Mg 2+ as activa- tor),orevento45°C(Mn 2+ as activator). This temperature stability is not surprising in view of the optimum growth of M. laminosus at 50–55 °C.ThedatainTable2were obtained after 60 min reaction time, the activities are given as the amplitude of the reversible photochemistry of the product, a-PEC, relative to that of the reaction at the optimum temperature, 45 °C. For longer reaction times (180 min), the temperature optimum is reduced to 37 °C (Mn 2+ ), the activity at 45 °C being 20% less. Apparently, the lyase has a high transient at 45 °C, but also becomes more rapidly inactivated than at 37 °C, probably by precipitation. Because of this inactivation, no experiments were carried out at T > 45 °C. At these temperatures, photo-oxidative side-reactions also become prominent (data not shown). In vivo, M. laminosus cells can prevent oxidation and protect the lyase from precipitation at considerably higher temperatures, up to 55 °C. Detergents. Although the His-tagged lyase as well as PecA are well water soluble at temperatures £ 37 °C, addition of mild detergents [e.g. 0.2–1% (v/v) Triton X-100, Nonidet P-40, Tween-20] was beneficial, doubling reaction speed. There was also another beneficial effect: when the Triton X-100 was present in the reaction system, the spontaneous addition of PCB to PecA forming adducts was reduced, thereby increasing the proportion of isomerization product, 3 1 -Cys-PVB-PecA (as demonstrated before [24], the PCB- PecA adduct cannot be transformed to a-PEC by His 6 -PecE and His 6 -PecF). Addition of Triton X-100 to the reaction mixture resulted in an absorption shift of the long wavelength band of PCB from 620 nm to 600 nm (Fig. 2), irrespective of the presence of the lyase (PecE/F) and the structural protein, PecA. This implies that the absorption change was due to the amphipathic property of Triton X-100. Possibly, Triton X-100 modifies the confor- mation of chromophore, to a form suitable for the PVB- PEC-lyase to act on, and unfavourable for PecA to bind spontaneously to PCB. Changes of the conformational equilibria of bile pigments have been reported in a variety of environments [27,28], including lipids [29]. It is possible that, by analogy, PecE/F also changes the conformation of the bilin in the course of the addition reaction (see below). After optimization of the enzyme reaction, the reconsti- tution is accelerated by  10-fold as compared with the original conditions [23], and the amount of the spontaneous addition product, 3 1 -Cys-PCB (k max ¼ 640 nm) is at the same time reduced. Absorption spectra and light-induced changes of typical reconstitution mixtures are shown in Fig. 3A and B, and those of a product purified by affinity chromatography in Fig. 3C. Note the relatively high absorption (580–600 nm) between the two major peaks, Fig. 1. The effect of different pH on the lyase/isomerase action of His 6 - PecE/F. Except for the pH, the reaction was carried out under opti- mized conditions (20 l M each PecA, PecE and PecF; 25 l M PCB, 5m M ME, 3 m M Mn 2+ , see Materials and methods for details). The product was assayed by the reversible photochemistry of the correct product, 3 1 -Cys-PVB-PecA (a-PEC) according to Zhao et al. [11]. Table 2. Temperature dependance of activity of His 6 -PecE and His 6 -PecF. All reactions were carried out in the optimized potassium phosphate/Tris buffer system (see text) in the presence of Mn 2+ (3 m M ) and 2-mercaptoethanol (5 m M ). The yield of photoactive a-PEC was assayed after 60 min. Temperature (°C) Relative activity (%) 20 20 30 43 37 67 45 100 Fig. 2. Interaction of PCB with Triton X-100. Addition of Triton X-100 resulted in a blue shift of the absorption of PCB from 620 nm to 600 nm, both in the presence of His 6 -PecA, His 6 -PecE and His 6 -PecF (A), and in the absence of these proteins (B). 4546 K H. Zhao et al. (Eur. J. Biochem. 269) Ó FEBS 2002 which is not observed when the reconstitution is carried out under catalysis of untagged PecE and PecF [24]. This absorption is seen only when His 6 -PecE was used, and in particular when Triton X-100 was added to the reaction mixture (Fig. 3B). It is lost when the reaction mixture is subjected to Ni 2+ chelating chromatography (Fig. 3C), which shows that it derives from free chromophore(s). Kinetics of enzymatic ligation/isomerization The kinetics of the formation of His-tagged a-PEC from PCB and His 6 -PecA, catalysed by His 6 -PecE and His 6 - PecF, follows the Michaelis–Menten equation for each substrate, i.e. PCB (Fig. 4A) and His 6 -PecA (Fig. 4B). The kinetic constants derived from these plots are summarized in Table 3. Like the only other phycobilin-lyase studied [30], the 3 1 -Cys84-PecA:PCB lyase is a rather slow enzyme (k cat ¼ 10 )4 )10 )5 Æs )1 ) with moderate affinity. The Line- weaver–Burk plots with respect to PCB, obtained at different concentrations of His 6 -PecA, intersect within the limits of error at a common point which seems not to be located on the x-axis. Such behaviour is typical for a sequential mechanism of the enzyme reaction with the two substrates bound one after the other [31,32]. Action of individual subunits, PecE and PecF As shown before, both lyase subunits, PecE and PecF, are necessary for the reconstitution of PCB and PecA to yield Fig. 3. His 6 -PecE/F catalysed ligation and isomerization of PCB with PecA. Photochemistry [before (solid line) and after (dashed line) saturating irradiation with 570 nm light) of the reconstitution mixture (PCB plus His 6 -PecA, His 6 -PecE and His 6 -PecF) under otherwise optimized conditions (see Fig. 1 and Materials and methods), in the absence (A) and presence (B) of Triton X-100. Note the relatively strong absorption at 580–600 nm (arrow) between the two product bands, which is increased in the presence of Triton X-100. It is lost after Ni 2+ chelating chromatography (C). Fig. 4. Enzyme kinetics of the PVB-PEC-lyase. Lineweaver–Burk plots of the ligation–isomerization reaction catalysed by His 6 -PecE and His 6 -PecF, for the two substrates, PCB (A) and His 6 -PecA (B). Other conditions were as described in Fig. 1. At different concentrations of His 6 -PecA, the corresponding linear fits do not intersect on the x-axis. Ó FEBS 2002 Enzymology of isomerizing phycoviolobilin lyase (Eur. J. Biochem. 269) 4547 the phycoviolobilin-bearing chromoprotein, a-PEC [23], and neither of the two subunits alone could catalyse the reconstitution effectively. In this reconstitution, the enzyme catalyses two reactions: the covalent binding of PCB to the apo-protein, and its transformation to bound PVB. It was therefore interesting to see if and how the two subunits PecE and PecF, which show a low degree of homology, function in the absence of the other. A careful inspection of the absorption changes (Fig. 5) indeed showed some subunit- specific residual activities: PecE applied alone, increases the ÔspontaneousÕ or auto(?) catalytic binding of PCB to PecA by 25%, yielding, however, only 3 1 -Cys84-PecA-PCB. In the presence of His 6 -PecF, this pure addition reaction of PCB to His 6 -PecA was decreased 15%. However, in this case a small amount of the ligation/isomerization product, His 6 -a-PECA, was formed (7% as compared to the maximal yield of His 6 -a-PecA in the presence of His 6 -PecE and His 6 -PecF). This may indicate that PecE is mainly responsible for binding the chromophore to the apoprotein, PecA, and PecF is mainly promoting the isomerization PCB to PVB. Interestingly, this model is supported by sequence comparison between the respective subunits of the two enzymes(Table4):Forthetwoorganismsforwhichthe sequences are known, there is a significantly higher homo- logy and Z-score for the E-subunits that for the F-subunits. If this functional distinction of the two subunits is correct, the question arises as to the function of the F-subunit of the phycocyanin lyases. Possibly, it acts as an isomerase as well in this case, but as one ensuring or chaperoning the isomerization of improperly bound chromophores, for example those having incorrect stereochemistry. It should be emphasized again, however, that the product of the ÔspontaneousÕ addition reaction (3 1 -Cys84-PecA- PCB) bearing the PCB chromophore, can not be isomerized to the PVB chromophore by the action of PecE and PecF, either alone or in combination. For some of the phyto- chromes, a sequential ligation reaction is discussed [14,33,34]. The isomerization may therefore proceed at an intermediate state. The concerted action of the lyase subunits is supported by a physical interaction between them. It had already been shownthatincaseofthePC-Cys-a84 lyase, CpcE and CpcF form a 1 : 1 complex [30]. Gel filtration experiments with PecE/F on HiPrep Sephacryl S-200 (Amersham Pharmacia Biotech AB) proved inconclusive. There were aggregates (50–60, 80–90, 120–140 kDa) observed in the mixture of the subunits, but both His 6 -PecE, and His 6 -PecF formed homo- oligomers (e.g. dimer, trimer, and tetramer), and the resolution was insufficient to clearly distinguish homo- from hetero-oligomers (data not shown). However, the formation of complexes between the subunits, PecE and PecF; is supported by the following experiments. In the first approach, His 6 -PecF was absorbed on a Ni 2+ chelating column in start buffer (0.5 M NaCl, 20 m M potassium Table 3. Enzymatic parameters for the ligation/isomerization of PCB to His 6 -PecA, catalysed by His 6 -PecE/F under optimum conditions (see text). Data were derived from the fits shown in Fig. 4. [PecE] and [PecF] were 8.6 l M ,sok cat ¼ v max /8.6 · 10 )6 . Substrate (S) Co-substrate (concentration) K S m (l M ) v max (n M Æs )1 ) k cat (s )1 ) PCB PecA (14 l M ) 2.4 0.65 0.76 · 10 )4 PecA PCB (19 l M ) 12 1.1 1.3 · 10 )4 PecA PCB (9.6 l M ) 14 0.97 1.1 · 10 )4 PecA PCB (4.8 l M ) 16 0.95 1.1 · 10 )4 Fig. 5. Effect of individual lyase–isomerase subunits on the reaction of PCB with PecA. (A) Compared to the spontaneous addition reaction (solid line), direct binding of PCB to His 6 -PecA without isomerization to 3 1 -Cys-PVB, is increased by 25% in the presence of His 6 -PecE (dotted line), but no His 6 -a-PecA (k max ¼ 570 nm) is formed. In the presence of His 6 -PecF (dashed line), the yield of PCB-His 6 -PecA is decreased by 15%, and a shoulder at 570 nm is clearly visible. It was identified as 3 1 -Cys-PVB-PecA and quantitated by its reversible pho- tochemistry (B). Large amounts of the ÔcorrectÕ ligation–isomerization product 3 1 -Cys-PVB-PecA were formed only in the presence of both PecE and PecF [solid line in (A)]. Table 4. Amino acid identities and Z-scores between the respective subunits of PCB-lyases and PVB-PEC-lyase from M. laminosus (Fischerella PCC 7603 [35,36], D. Wu, J P. Zhu, H. Scheer, K H. Zhao, unpublished results, GenBank AF506031) and Anabena sp. PCC7120 [37–40]. Amino acid identities [%] (Z-score) for comparison of Organism CpcE/PecE CpcF/PecF M. laminosus 47.5 (311) 34.5 (212) Anabena spec 46.7 (387) 32 (286) 4548 K H. Zhao et al. (Eur. J. Biochem. 269) Ó FEBS 2002 phosphate, pH 7.2), and then the same amount of untagged PecE [24] dissolved in the start buffer, was applied to the preloaded column. It was then first washed with the start buffer, and subsequently with stripping buffer containing EDTA (100 m M )andNaCl(0.5 M ). In the first washing fractions with start buffer, 70% of the PecE (as judged by SDS/PAGE) was eluted. The remaining 30% of the PecE stayed on the column during further washing, in spite of it lacking a His-tag, and was eluted only with the stripping buffer together with the majority of the His-tagged PecF. Independent support for the formation of complexes between PecE and PecF comes also from reversible denaturation experiments (Table 5): Denaturation in 8 M urea of the individual subunits, PecE or PecF, is largely irreversible: if they are mixed together after dialysing out the urea separately, they show only little activity. This is also true if either individually treated PecE is mixed with native PecF, or vice versa. By contrast, if the two subunits are mixed in the denatured state and then the urea is dialysed out from the mixture, the resulting product shows full activity. CONCLUSIONS Cofactor requirements and enzyme kinetics of PVB-PEC- lyase from M. laminosus have been studied. The novel, dual- action enzyme is responsible for the attachment and isomerization of phycocyanobilin to PecA, the a-subunit of phycoerythrocyanin. Mercaptoethanol and the divalent metals, Mg 2+ or Mn 2+ , were required, and the reaction was aided by the detergent Triton X-100. The speed of the reaction and the purity of the products was improved by careful adjustment of the buffer, balancing in particular the conflicting effects of potassium phosphate buffer, which inhibits protein precipitation, but at the same time binds the required metal. These improvements will provide a basis for the preparative reconstitution of the individually or jointly modified reaction partners, viz. the structural protein PecA and the substrate chromophore, PCB. Kinetic experiments showed the enzyme to be rather slow, comparable to a related mono-functional PCB- phycocyanin lyase [30]. Furthermore, they indicated that the reconstitution reaction proceeds by a sequential mech- anism, which has the characteristics that the enzyme reaction requires all of the substrates to be present before any product is released. This is consistent with HPLC results detecting no chromophore other than the substrate PCB in the reaction mixture [23]. Moreover, there is evidence, that PecE is responsible for chromophore binding, and PecF for the isomerization. However, although PCB does bind covalently to His 6 -tagged PecA to form PCB-His 6 -tagged PecA, the latter is no substrate of the enzyme: it could not be transformed to PVB-His 6 -tagged PecA (i.e. His 6 -tagged a-PecA) under catalysis of PecE and/or PecF. By using a combination of untagged and His-tagged subunits, evidence was obtained for the interaction between PecE and PecF. Experiments of this type are expected to guide the way to ternary and quaternary complexes of the unusual enzyme. The ligation mechanism of the chromophores to phyco- bilin and phytochrome apoproteins still remains largely unknown. It is hoped that other isomerizing lyases leading to biliproteins will be characterized in the future, in particular those yielding chromophores with a D2,3-double bond (phycourobilin, several cryptophytan proteins). ACKNOWLEDGEMENTS The laboratory of K.H.Z. is supported by Natural Science Foundation of China (project number 39770175). K.H.Z. is grateful to the DAAD, Bonn, Germany for a fellowship, and to the Alexander von Humboldt Foundation, Bonn, Germany for donation of a microcentrifuge subsequent to a postdoctoral fellowship. The laboratory of H.S. is supported by Deutsche Forschungsgemeinschaft (SFB 533, TPA1). REFERENCES 1. Glazer, A.N. (1994) Adaptive variations in phycobilisome struc- ture. Adv. Mol. Cell Biol. 10, 119–149. 2. Grossman, A.R., Schaefer, M.R., Chiang, G.G. & Collier, J.L. (1993) The phycobilisome, a light-harvesting complex responsive to environmental conditions. Microbiolog. Rev. 57, 725–749. 3. Schluchter, W.M. & Glazer, A.N. (1999) Biosynthesis of phyco- biliproteins in Cyanobacteria In: The Phototrophic Prokaryotes (Peschek, G., Lo ¨ ffelhardt, W. & Schmetterer, G., eds), pp. 83–95. Kluwer Academic/ Plenum Press, New York. 4. Beale, S.I. (1994) Biosynthesis of cyanobacterial tetrapyrrole pig- ments: hemes, chlorophylls and phycobilins. In The Molecular Biology of Cyanobacteria (Bryant, D.A., ed), pp. 519–558. Kluwer, Dordrecht. 5. Ru ¨ diger, W. & Thu ¨ mmler, F. (1991) Phytochrom, the visual pigment of plants. Angew. Chem. International 30, 1216–1228. 6. Smith, H. (2000) Phytochromes and light signal perception by plants–anemergingsynthesis.Nature 407, 585–591. 7. Sineshchekov, V.A. (1999) Phytochromes: molecular structure, photoreceptor process and physiological function. In Concepts in Photobiology (Singhal, G.S., ed.), pp. 755–795. Narosa, New Delhi. 8. Hughes, J. & Lamparter, T. (1999) Prokaryotes and phytochrome. The connection to chromophores and signaling. Plant Physiol. 121, 1059–1068. 9. McDowell, M.T. & Lagarias, J.C. (2002) Analysis and recon- stitution of phytochromes. In Heme, Chlorophyll, and Bilins (Smith, A. & Witty, M., eds), pp. 293–309. Humana Press, Totawa, USA. Table 5. Activites of the isomerizing PVB-PEC-lyase after reversible denaturation with 8 M urea, determined by the photochemical activity DDA ([11]) of the product mixture, relative to that of a-PEC produced in the enzymatic reaction under optimized conditions. The mixtures shown on the left were reacted for 60 min with PCB. PecE and PecF denote the native subunits of the enzyme as isolated, PecE ren and PecE ren the subunits treated individually first with 8 M urea, and then dialysed against potassium phosphate buffer. In the last experiment (PecE + PecE) ren denotes a mixture of PecE and PecF subjected under other- wise identical conditions to the denaturation/dialysis cycle. Lyase/substrate fraction Relative activity (%) PecA + PecE 0 PecA + PecF 7 PecA + PecE + PecF 100 PecA + PecE ren + PecF 3 PecA + PecE + PecF ren 5 PecA + PecE ren + PecF ren 3 PecA + (PecE + PecF) ren 112 Ó FEBS 2002 Enzymology of isomerizing phycoviolobilin lyase (Eur. J. Biochem. 269) 4549 10. Zhao, K.H. & Scheer, H. (1995) Type I and type II rever- sible photochemistry of phycoerythrocyanin a-subunit from Mastigcoladus laminosus both involve Z, E isomerization of phycoviolobilin chromophore and are controlled by sulfhydryls in apoprotein. Biochim. Biophys. Acta 1228, 244–253. 11. Zhao, K.H., Haessner, R., Cmiel, E. & Scheer, H. (1995) Type I reversible of phycoerythrocyanin involves Z/E-isomerization of a-84 phycoviolobilin chromophore. Biochim. Biophys. Acta 1228, 235–243. 12. Willows, R.D., Mayer, S.M., Foulk, M.S., DeLong, A., Hanson, K., Chory, J. & Beale, S.I. (2000) Phytobilin biosynthesis: the Synechocystis sp. PCC 6803 heme oxygenase-encoding ho1 gene complements a phytochrome-deficient Arabidopsis thaliana hy1 mutant. Plant Mol. Biol. 43, 113–120. 13. Frankenberg, N., Mukougawa, K., Kohchi, T. & Lagarias, J.C. (2001) Functional genomic analysis of the hy2 family of ferre- doxin-dependent bilin reductases from oxygenic photosynthetic organisms. Plant Cell 13, 965–978. 14. Wu, S H. & Lagarias, J.C. (2000) Defining the bilin lyase domain: lessons from the extended phytochrome superfamily. Biochemistry 39, 13487–13495. 15. Beltrame, P.L., Monti, D., Sala, R. & Manitto, P. (1985) Addition of 2-mercaptopropionylglycine to biliverdin – a kinetic and ther- modynamic study. Gaz. Chim. Ital. 115, 223–226. 16. Manitto, P. & Monti, D. (1979) Addition of sulphydryl groups to biliverdin. Experientia 35, 1418–1420. 17. Fairchild, C.D. & Glazer, A.N. (1994) Nonenzymatic bilin addi- tion to the gamma subunit of an apophycoerythrin. J. Biol. Chem. 269, 28988–28996. 18. Arciero, D.M., Dallas, J.L. & Glazer, A.N. (1988) In vitro attachment of bilins to apophycocyanin. 2. Determination of the structures of tryptic bilin peptides derived from the phycocyano- bilin adduct. J. Biol. Chem. 263, 18350–18357. 19. Arciero, D.M., Bryant, D.A. & Glazer, A.N. (1988) In vitro attachment of Bilins to Apophycocyanin. 1. Specific covalent adduct formation at cysteinyl residues involved in phycocyano- bilin binding in C-phycocyanin. J. Biol. Chem. 263, 18343–18349. 20. Fairchild, C.D., Zhao, J., Zhou, J., Colson, S.E., Bryant, D.A. & Glazer, A.N. (1992) Phycocyanin a-subunit phycocyanobilin lyase. Proc. Natl. Acad. Sci. USA 89, 7017–7021. 21. Fairchild, C.D. & Glazer, A.N. (1991) Bilin addition to the a subunit of apophycocyanin. In Photosynthetic Procaryotes (Fuller, R.C., ed.). Amherst, USA. 22. Schluchter, W.M. & Bryant, D.A. (2001) Analysis and recon- stitution of phycobiliproteins. methods for the characterization of bilin attachment reactions. In Heme, Chloropyll and Bilins,(Smith, A. & Witty, M., eds), pp. 311–334. Humana Press, Totawa, USA. 23. Storf, M., Parbel, A., Meyer, M., Strohmann, B., Scheer, H., Deng, M., Zheng, M., Zhou, M. & Zhao, K. (2001) Chromophore attachment to biliproteins. Specificity of PecE/ PecF, a lyase/isomerase for the photoactive 3 1 -Cys-a ´ 84-phycovi- olobilin chromophore of phycoerythrocyanin. Biochemistry 40, 12444–12456. 24. 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 PecE and PecF. FEBS Lett. 469, 9–13. 25. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriopharge T4. Nature 227, 680–685. 26. Scheer, H., Linsenmeier, U. & Krauss, C. (1977) Chemical and photochemical oxygenation of a phytochrome P r chromophore model pigment to purpurins. Hoppe-Seyler’s Z. Physiol. Chem. 358, 185–196. 27. Falk, H. (1989) The Chemistry of Linear Oligopyrroles and Bile Pigments. Springer, Wien, New York. 28. Braslavsky, S.E., Holzwarth, A.R. & Schaffner, K. (1983) Solu- tion confirmations, photophysics, and photochemistry of bili- pigments – bilirubin and biliverdin dimethylesters and related linear tetrapyrroles. Angew. Chem. International 22, 656–674. 29. Tegmo-Larsson, I M., Braslavsky, S.E., Culshaw, S., Ellul, R.E., Nicolau, C. & Schaffner, K. (1981) Phytochrome models. 6. Conformation control by membrane of biliverdin dimethylester incorporated into lipid vesicles. J. Am. Chem. Soc. 103, 7152–7158. 30. Fairchild, C.D. & Glazer, A.N. (1994) Oligomeric Structure, enzyme kinetics, and substrate specificity of the phycocyanin alpha subunit phycocyanobilin lyase. J. Biol. Chem. 269, 8686–8694. 31. Hammes, G.G. (1982) Enzyme Catalysis and Regulation. Aca- demic Press, New York. 32. King, E.L. & Altman, C. (1956) A schematic method of deriving the rate laws for enzyme-catalyzed reactions. J. Phys. Chem. 60, 1375–1378. 33. Davis, S.J., Vener, A.V. & Vierstra, R.D. (1999) Bacterio- phytochromes: phytochrome-like photoreceptors from non- photosynthetic eubacteria. Science 286, 2517–2520. 34. Lamparter, T., Esteban, B. & Hughes, J. (2001) Phytochrome Cph1 from the cyanobacterium Synechocystis PCC6803. Puri- fication, assembly, and quaternary structure. Eur. J. Biochem. 268, 4720–4730. 35. Kufer, W., Ho ¨ gner, A., Eberlein, M., Mayer, K., Buchner, A. & Gottschalk, L. (1991) Structure and molecular evolution of the gene cluster encoding proteins of the rod substructure of the phycobilisome from the cyanobacterium Mastigocadus laminosus. Gene Bank, M75599. 36. Glauser, M., Stirewalt, V.L., Bryant, D.A., Sidler, W. & Zuber, H. (1992) Structure of genes encoding rod-core linker polypeptides of Mastigocladus laminosus phycobilisomes and functional aspects of the phycobiliprotein/linker–polypeptide interactions. Eur. J. Bio- chem. 205, 927–937. 37. Belknap, W.R. & Haselkorn, R. (1987) Cloning and light reg- ulation of expression of the phycocyanin operon of the cyano- bacterium Anabaena. EMBO J. 6, 871–884. 38. Swanson, R.V. Lorimier de, R. & Glazer, A.N. (1992) Genes encoding the phycobilisome rod substructure are clustered on the Anabaena chromosome: Characterization of the phycoery- throcyanin operon. J. Bacteriol. 174, 2640–2647. 39. Bryant, D.A., Stirewalt, V.L., Glauser, M., Frank, G., Sidler, W. & Zuber, H. (1991) A small multigene family encodes the rod-core linker polypeptides of Anabaena Sp PCC7120 phycobilisomes. Gene 107, 91–99. 40. Kaneko, T., Nakamura, Y., Wolk, C.P., Kuritz, T., Sasamoto, S., Watanabe, A., Iriguchi, M., Ishikawa, A., Kawashima, K., Kimura, T., Kishida, Y., Kohara, M., Matsumoto, M., Matsuno, A., Muraki, A., Nakazaki, N., Siumpo, S., Sugimoto, M., Taka- zawa,M.,Yamada,M.,Yasuda,M.&Tabata,S.(2001)Com- plete genomic sequence of the filamentous nitrogen-fixing cyanobacterium Anabaena sp. strain PCC 7120. DNA Res. 8,205– 213. 41.Bishop,J.E.,Rapoport,H.,Klotz,A.V.,Chan,C.F.,Glazer, A.N., Fu ¨ glistaller, P. & Zuber, H. (1987) Chromopeptides from phycoerythrocyanin. Structure and linkage of the three bilin groups. J. Am. Chem. Soc. 109, 875–881. 4550 K H. Zhao et al. (Eur. J. Biochem. 269) Ó FEBS 2002 . Characterization of phycoviolobilin phycoerythrocyanin-a84-cystein- lyase-(isomerizing) from Mastigocladus laminosus Kai-Hong Zhao 1 ,. Interaction of PCB with Triton X-100. Addition of Triton X-100 resulted in a blue shift of the absorption of PCB from 620 nm to 600 nm, both in the presence of