132,173-Cyclopheophorbide B Enol As a Catabolite of Chlorophyll B In

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

FEBS Letters 587 (2013) 2578–2583

journal homepage: www.FEBSLetters.org

132,173-Cyclopheophorbide b enol as a catabolite of chlorophyll b in phycophagy by protists ⇑ Yuichiro Kashiyama a,b,c, , Akiko Yokoyama d, Takashi Shiratori d, Isao Inouye d, Yusuke Kinoshita a, ⇑ Tadashi Mizoguchi a, Hitoshi Tamiaki a,

a Graduate School of Life Sciences, Ritsumeikan University, Kusatsu, Shiga 525-8577, Japan b Department of Environmental and Biological Chemistry, Fukui University of Technology, Fukui, Fukui 910-8505, Japan c Precursory Research for Embryonic Science and Technology (PRESTO), Japan Science and Technology Agency (JST), Chiyoda, Tokyo 153-8902, Japan d Graduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan

article info abstract

Article history: Both 132,173-cyclopheophorbide a and b enols were produced along with ingestion of green Received 9 May 2013 microalgae containing chlorophylls a and b by a centrohelid protist (phycophagy). The results sug- Revised 24 June 2013 gest that chlorophyll b as well as chlorophyll a were directly degraded to colored yet non-phototoxic Accepted 25 June 2013 catabolites in the protistan phycophagic process. Such a simple process by the predators makes a Available online 4 July 2013 contrast to the much sophisticated chlorophyll degradation process of land plants and some algae, where phototoxicity of chlorophylls was cancelled through the multiple enzymatic steps resulting in Edited by Richard Cogdell colorless and non-phototoxic catabolites. Ó 2013 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Keywords: Aquatic ecosystem Chlorophyll catabolism Cyclopheophorbide enol Microalga Phycophagy Protist

1. Introduction colorless, non-fluorescent catabolites, hence abolishing the phototoxicity of Chls in the plant tissues. Chlorophylls (Chls) are essential components of the photosyn- Recently, a second degradation process of Chl-a (4) has been thetic apparatus, and thus significant molecules for living organ- proposed [2], which is conducted in cells of phycophagic protists isms and environments. Whereas the biosynthesis of Chls is (protists feeding on microalgae) following their ingestion of algal relatively well understood, studies on their biodegradative pro- diets. The resulting product was 132,173-cyclopheophorbide a enol cesses are still limited. In land plants, Chl-a (4 in Fig. 1) is decom- (cPPB-aE; 11), which is a green colored but non-fluorescent pig- posed step-by-step, where pheophorbide a oxygenase (PAO) plays ment. Due to its non-phototoxicity, cPPB-aE represents another a central role that cleaves the tetrapyrrole macrocycle generating a detoxified catabolite of Chl-a. The catabolite is ubiquitously de- linear tetrapyrrolic catabolite, a so-called PAO-pathway [1]. The tected in oceans and lakes where large amounts of cPPB-aE are catabolism through the PAO-pathway eventually results in accumulated as the most abundant Chl-a-derived degradation product [2,3]. Therefore, cPPB-aE metabolism is a widely distrib- uted degradation process of Chl-a in aquatic ecosystems. Abbreviations: APCI, atmospheric pressure chemical ionization; Chl, chlorophyll; Little has been understood about the degradation of other Chl cPPB-aE, 132,173-cyclopheophorbide a enol; cPPB-bE, 132,173-cyclopheophorbide b molecules along with protistan phycophagy, including Chl-b (1) enol; PAO, pheophorbide a oxygenase; PDA, photodiode array; Phe, pheophytin; in green algae. Green algae include diverse forms of aquatic micro- PPB, pheophorbide; pPhe, pyropheophytin; pPPB, pyropheophorbide; (R/S)-hCPLs, and macroalgae as well as land plants (embryophytes) and com- (132R)- and (132S)-hydroxychlorophyllones; TOF, time-of-flight monly produce Chl-b as an accessory pigment in photosynthetic ⇑ Corresponding authors at: Graduate School of Life Sciences, Ritsumeikan University, Kusatsu, Shiga 525-8577, Japan. Fax: +81 776297891 (Y. Kashiyama), antenna. Chl-b in land plants is known to be decomposed through +81 775612659 (H. Tamiaki). the PAO-pathway, where Chl-b is first converted into Chl-a in a E-mail addresses: [email protected] (Y. Kashiyama), [email protected] two-step enzymatic process, and further decomposed [1]. Such a (H. Tamiaki).

0014-5793/$36.00 Ó 2013 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.febslet.2013.06.036 Y. Kashiyama et al. / FEBS Letters 587 (2013) 2578–2583 2579

O R1

N N Chl-b (1): M = Mg; R = COOMe N N Chl-a (4): M = Mg; R = COOMe NH N M Phe-b (2): M = 2H; R = COOMe M Phe-a (5): M = 2H; R = COOMe N N pPhe-b (3): M = 2H; R = H N N pPhe-a (6): M = 2H; R = H N HN

R O R O O O a 7 1 2 O O O O OR2 pPPB- ( ): R = Me; R = H pPPB-b (8): R1 = CHO; R2 = H Methyl pPPB-b (9): R1 = CHO; R2 = Me

O O O

NH N NH N NH N NH N NH N

N HN N HN N HN N HN N HN

OH OH O H O OH O OH O O O O O

cPPB-bE (10) cPPB-aE (11) Diketone form of 10 (12) (R/S)-hCPL-b (13) (R/S)-hCPL-a (14)

Fig. 1. Chemical structures of natural chlorophylls (Chls) and their derivatives. controlled degradation of the phototoxic Chl-b is expected also in 2.2. Feeding experiments aquatic ecosystems, because green ‘‘plants’’ are important components particularly in fresh water environments (i.e., microalgae Feeding experiments were designed for the above maintained including such familiar genera as Chlamydomonas, Chlorella, Closte- centrohelid SRT127, to which an aliquot of unialgal culture of Pyra- rium, Pediastrum, Spyrogyra, Staurastrum, and Volvox). For example, mimonas sp. (DA140) was added periodically every 3 to 10 days. a green alga Chlorella protothecoides was reported to conduct a The mixture was incubated at 20 °C in artificial seawater (Wako) PAO-like chlorophyll degradation, where both Chl-a and Chl-b enriched with Daigo’s IMK medium under RGB LED light at were directly converted into linear-tetrapyrrolic catabolites [4]. 16 lmol photon m2 s1, and the whole culture were filtrated We here report quantitatively significant occurrences of a Chl-b through sterile GF/F filters (Whatman, 47 mm /; GE Healthcare, derivative that is analogous to cPPB-aE, along with the consump- Buckinghamshire, UK). Total incubation periods were typically tion of green microalgae by phycophagic protists, hence proposing 20 days. Cultured cells of SRT127 were isolated from the above a direct degradation pathway of Chl-b (1)to132,173-cyclopheo- mixture containing DA140 by micropipetting, co-cultured with phorbide b enol (cPPB-bE; 10), without conversion to Chl-a in the cryptomonad (OR1), then incubated at 20 °C in filtration-ster- advance. ilized natural seawater enriched with an ESM medium for 5 days before filtration through GF/F filters. The GF/F filters were immedi- 2. Materials and methods ately frozen and stored in 20 °C before pigment extraction.

2.1. Strains of protists 2.3. Sample preparation

A strain of a colorless centrohelid (SRT127; undescribed) was Each wet filter was extracted three times by ultrasonication in isolated by micropipetting from a seawater sample collected at a acetone (‘‘dioxin analysis’’ grade; Wako Pure Chemicals) for wharf of Tokyo Bay, Tokyo, Japan (35° 370 500 N, 139° 460 2300 W) 5 min at 0 °C in the dark. An aliquot of ethanol (‘‘dioxin analysis’’ on July 30, 2011. SRT127 was co-cultured with a strain of green grade, Wako Pure Chemicals) was added to the combined extracts, alga, Pyramimonas sp. (DA140), which was also isolated from the then dried in vacuo in the dark. The residue was dissolved in ani- Ò same location. Unialgal DA140 cultures as well as co-cultures of sole (‘‘RegentPlus ’’ grade; Sigma–Aldrich, St. Louis, USA) just be- SRT127 with DA140 were maintained at 18 °C in filtration- fore analysis. All of the above operations were performed under sterilized natural seawater enriched with Daigo’s IMK medium argon atmosphere using a glove box. for marine microalgae (Wako Pure Chemicals, Osaka, Japan) under fluorescent lights at 12 lmol photon m2 s1. A strain of a crypto- 2.4. HPLC analysis/isolation of pigments monad alga Chroomonas sp. (OR1) was isolated by micropipetting from a seawater sample collected at Ooarai Sun Beach, Ooarai Iba- Analytical HPLC was performed using a Shimadzu Prominence raki, Japan (36° 180 1000 N, 140° 340 500 W) on June 23, 2010, which liquid chromatograph system, comprised of a CBM-20A communi- was maintained at 18 °C in filtration-sterilized natural seawater cations bus module, a DGU-20A5R degasser, three LC-20AD pumps enriched with an ESM medium [5] under fluorescent lights (Cool constituting a ternary pumping system, an SIL-20AC auto sampler, White Fluorescent Lamp; Mitsubishi OSRAM, Yokohama, Japan) a CTO-20AC column oven, an SPD-M20Avp photodiode array (PDA) at 12 lmol photon m2 s1. detector, and an FRC-10 automated fraction collector (Kyoto, 2580 Y. Kashiyama et al. / FEBS Letters 587 (2013) 2578–2583

(R)-hCPL-a (13) (S)-hCPL-a (13) cPPB-aE (11) Chl-a (4) Phe-b (2) Phe-a (5) pPhe-a (6) b 1 a’ b 3 A pPPB-b (8) pPPB-a (7) Chl- ( ) Chl- pPhe- ( ) 400

500

600

700

5 10 15 20 25 B cPPB-bE (10) 400

500

600

700 C 5 10 15 20 25 400

500

600

700

5 10 15 20 25 x y z (carotenoid) Chl-b’ D 400 Wavelength (nm) 500

600

700

5 10 15 20 25 Fraction I Fraction II E 400

500

600

700

5 10 15 20 25 F 400

500

600

700

5 10 15 20 25 Retention time (min)

Fig. 2. Three-dimensional HPLC chromatograms. (A) An authentic sample mixture. (B) Crude products containing synthetic cPPB-bE. (C) Extract of Pyramimonas sp. (DA140). (D) Extract of co-culture of centrohelid (SRT127) with DA140. (E) Extract of cryptomonad (OR1). (F) Extract of co-culture of SRT127 with OR1.

Japan). The system was coupled to a personal computer configured pressure, 1600 hPa; drying gas temperature, 200 °C; vaporizer to run Shimadzu LC Solution software. Reverse-phase HPLC was temperature, 350 °C; capillary voltage (positive), 4000 V; corona performed under the following conditions: column, ZORBAX current, 216 nA. Eclipse Plus C18 (Rapid Resolution HT, 4.6 / 30 mm, 1.8 lm sil- ica particle size; Agilent Technologies, Santa Clara, USA); eluent, 2.6. Authentic samples a ternary gradient program summarized in Supplementary Fig. S1; flow rate, 1.00 ml min1; range of detection wavelength Chl-a (4), pheophytin a (Phe-a; 5), pyropheophytin a (pPhe-a; by PDA, 300–800 nm. All solvents used for the analytical HPLC mo- 6), pyropheophorbide a (pPPB-a; 7), cPPB-aE(11), and (132R)- bile phases were of ‘‘HPLC grade’’ quality, and were purchased and (132S)-hydroxychlorophyllones a ((R/S)-hCPLs-a; 14) were from Nacalai Tesque (Kyoto). prepared [2]. Chl-b (1) was prepared from a spinach extract and purified by preparative HPLC. Chl-b derivatives including Phe-b 2.5. APCI-TOF-MS analysis (2), pPhe-b (3), and methyl pPPB-b (9) were derived using the identical methods as the Chl-a derivatives described here and previ- Accurate molecular mass spectra of isolated pigments were ously [2]. Synthetic procedures were performed under argon or determined using a time-of-flight mass spectrometer (TOF-MS; nitrogen atmosphere in the dark. Preparative HPLC (Cosmosil micrOTOF-II; Bruker Daltonik, Bremen, Germany) coupled with 5C18-AR-II, 10 / 250 mm) was performed on a Shimadzu liquid an atmospheric pressure chemical ionization (APCI). The APCI con- chromatograph system comprised of an SCL-10Avp system conditions were set as follows: drying gas flow, 3.0 L min1; nebulizer troller, an LC-20AT pump and an SPD-M10Avp PDA. cPPB-bE(10) Y. Kashiyama et al. / FEBS Letters 587 (2013) 2578–2583 2581

A 1 C 1 b 2 Phe- ( ) Chl-b (1) S/R x ( )- Phe-b (2) 0.8 y hCPL-b (13) 0.8 z (cPPB-bE; 10) 0.2

0.6 0.6 0.1

0.4 0.4 0 Absorbance (a.u.) Absorbance (a.u.) 640 660 680

0.2 0.2

0 0 300 400 500 600 700 800 300400500600700800 Wavelength (nm) Wavelength (nm)

Soret Qy Soret Qy

B 1 Phe- a (5) D 1 Chl-a (4) 0.4 (S)-hCPL-a (14) Phe-a (5) (R)-hCPL-a (14) cPPB-aE (11) 0.8 0.8 0.2

0.6 0.6 0 640 660 680

0.4 0.4 Absorbance (a.u.) Absorbance (a.u.)

0.2 0.2

0 0 300 400 500 600 700 800 300400500600700800 Wavelength (nm) Wavelength (nm)

Soret Qy Soret Qy

Fig. 3. PDA on-line absorption spectra of Chl derivatives on HPLC analysis of the culture extracts. (A) Phe-b and peaks x and y (see Fig. 2D). (B) Phe-a and (S/R)-hCPL-a. (C) Chl-b, Phe-b, and peak z (see Fig. 2D). (D) Chl-a, Phe-a, and cPPB-aE. All the spectra were normalized at the most intense peak.

Chl-a (19.2)

Chl-b cPPB-aE (17.7) (S)-hCPL-a (10.1) (16.0)

(R)-hCPL-a (R/S)-hCPL-b (8.1, 8.3) (10.3) b x y cPPB- E Chl-b’ (14.2) (18.5) z Chl-a’ (20.0) Wavelength (nm) 640 a 650 Phe- 660 (23.8) 670 680 690 5 10 15 20 25 Retention time (min)

Fig. 4. Chromatographic peaks of the Chl derivatives. Numbers in parenthesis indicate retention time of the peak tops. 2582 Y. Kashiyama et al. / FEBS Letters 587 (2013) 2578–2583 was prepared from methyl pPPB-b (9) according to reported proce- The retention time of peak z was identical to that of the major com- dures for the cPPB-aE synthesis [2,6]. A solution of methyl pPPB-b ponent of the crude synthetic sample containing cPPB-bE, also sup- (3.1 mg, 5.5 lmol) in distilled THF (0.6 ml) was stirred for 3 min porting the above identification (Fig. 2B and D). under nitrogen atmosphere in the dark, to which commercially These assignments are consistent with chromatographic reten- available sodium bis(trimethylsilyl)amide ((Me3Si)2NNa, 25 ll, tion patterns relative to other Chl-b derivatives, which were simi- 38% in THF ca. 1.9 M, Tokyo Chem. Ind., Tokyo) was added and fur- lar to those of the analogous Chl-a derivatives (Fig. 4). The elution ther stirred for 30 min. After the color of the reaction mixture had pattern of peaks x and y (assigned as (R/S)-hCPLs-b) resembles the changed from green to dark orange, the solution was mixed into a double peak of (R/S)-hCPLs-a that elutes relatively quickly in the nitrogen-purged mixture of an aqueous saturated solution of so- chromatogram. Peak z (assigned as cPPB-bE) elutes ca. 3 min ear- dium dihydrogen phosphate (NaH2PO4, 2.0 ml) and dichlorometh- lier than Chl-b, which corresponds to the observation of cPPB-aE ane (8.0 ml), and stirred vigorously until the mixture became eluting ca. 3 min earlier than Chl-a. bright orange. The separated organic phase was washed with In addition, the comparison of online absorption spectra sup- water, dried over sodium sulfate and evaporated under a reduced ports the above assignments. The spectra of peaks x and y are very 1 pressure. The crude product (see its H NMR spectrum in Supple- similar to that of Phe-b, but Qy maxima of peaks x and y are slightly mentary Fig. S2) was not further purified but directly identified red- and blue-shifted, respectively, from that of Phe-b (Fig. 3A). by APCI-TOF-MS, where the primary mass peak at m/z = 531.2391 Such relationships in the Qy maxima closely resemble those be- exactly matched the calculated exact mass of MH+ of cPPB-bE tween (R/S)-hCPLs-a and Phe-a (Fig. 3B). The Soret band of peak z

(m/z = 531.2391, C33H30N4O3). was broadened more than those of Chl-b and Phe-b, extended to- ward the red side, and split into three peaks (365, 447, and 3. Results and discussion 474 nm) (Fig. 3C), which is comparable to the similarly broad Soret band of cPPB-aE that has two split peaks (359 and 423 nm) and a 3.1. Chlorophyll b-derived compounds shoulder (ca. 448 nm) (Fig. 3D). The Qy band of peak z is largely red-shifted relative to those of Chl-b and Phe-b (>20 nm) The PDA chromatogram of an extract from the unialgal culture (Fig. 3C), which is the same trend as observed for cPPB-aE relative of Pyramimonas sp. (DA140) obtained by the analytical HPLC shows to Chl-a and Phe-a (Fig. 3D). the presence of Chl-b (1) and Chl-a (4) as major components (Fig. 2C). Minor components of their 132-epimers, Chl-a0 and Chl- 3.3. Chl-b catabolism by phycophagic protists b0 would be attributable to artifacts during the sample preparation. cPPB-aE(11) was absent in the DA140 unialgal culture. On the Co-occurrence of cPPB-bE and cPPB-aE in the culture containing other hand, the chromatogram of an extract from the co-culture centrohelid with the Chl-a/Chl-b producing green alga demon- of the centrohelid (SRT127) and DA140 shows the presence of strated that cPPB-bE should have been produced by the centrohelid cPPB-aE as a major pigment, whereas Chl-a is virtually absent along with the phycophagic process on the alga in the same (Fig. 2D). Additionally, (R/S)-hCPLs-a (14) as well as Chl-b, Phe-b manner as previously reported for cPPB-aE [2]: the presumable (2), pPhe-b (3), Phe-a (5), and pPhe-a (6) were visible as unique enzymatic process(es) conducting aldol-like condensation of 2 3 components that were not present/prominent in the DA140 unial- C-13 and C-17 to form cPPB-aE from Chl-a. Thus, the centrohelid gal culture. seems to catabolize Chl-b directly into cPPB-bE without The centrohelid co-culture contains three other compounds converting it to Chl-a as reported in the degradation of Chl-b by (peaks x, y, and z) that were absent in the DA140 unialgal culture land plants [1]. and unidentified by the authentic samples (Fig. 2A). On-line spec- The bio-transformation of fluorescent Chl-b to non-fluorescent tra of these three peaks are shown in Fig. 3. These compounds are cPPB-bE(Supplementary Fig. S3) would be due to the detoxifica- likely to have been derived from Chl-b of the dietary green algae tion of photoactive Chl-b readily producing singlet oxygen. The re- Pyramimonas sp. because they were otherwise absent when the sult may imply that a single enzyme can take different chlorophylls centrohelid was co-cultured with a cryptomonad OR1 producing as its substrate. Such a rather prompt detoxification mechanism no Chl-b (Fig. 2E/F). Tentatively, we assigned these compounds to makes a contrast to the sophisticated PAO-pathway that rigorously be Chl-b derivatives, (R/S)-hCPLs-b (13) and cPPB-bE(10), which detoxifies chlorophylls by degrading into unconjugated hence non- were produced from Chl-b similarly as in the degradation of Chl- fluorescent linear tetrapyrroles: the PAO-pathway involves at least a (1)to(R/S)-hCPLs-a (14) and cPPB-aE(11). four enzymes [1] for fully detoxifying chlorophylls. On the other hand, the implied cPPB-aE/bE process is a more or less a first-aid 3.2. Assignment of chemical structures treatment since they can be reverted to phototoxic derivatives. In fact, (R/S)-hCPLs-b and (R/S)-hCPLs-a were also observed in To test the above assumption, two fractions containing peaks x these centrohelid cultures as major components. Since (R/S)- and y (fraction I) and peak z (fraction II) were analyzed by APCI- hCPLs-a were fluorescent [2] and phototoxic, it is not certain that TOF-MS. The observed peaks at m/z = 547.2354 for fraction I and the oxidized forms of cPPB-bE and cPPB-aE were actively produced at m/z = 531.2398 for fraction II definitely support the assign- by the centrohelid. Abiotic oxidations of cPPB-aE and cPPB-bE ments: peaks x and y as (R/S)-hCPLs-b (13;C33H30N4O4; exact mass would occur in the aged culture. We thus propose that Chl-b in of MH+: 547.2360) and peak z as cPPB-bE(10) or its diketone form the dietary algae should be actively catabolized into cPPB-bE, per- + (12)(C33H30N4O3; exact mass of MH : 531.2391), respectively haps by diverse phycophagic protists as reported for cPPB-aE. (Supplementary Table S1). These molecules could be derived through an aldol-like condensation from Chl-b, similarly as in the Acknowledgments transformation of Chl-a to cPPB-aE [7] and its oxidation to (R/S)- hCPLs-a [6]. Because no diketone form for the isomer of cPPB-aE This study was supported in part by PRESTO of the Japan has been reported, we assume that the enol form 10 of cPPB-bE Science and Technology Agency to Y.K., by Grants-in-Aid for is the more likely isomer for the present case. The presence of an Scientific Research awarded to Y.K. (23870028), A.Y. (20570081), enol function in a periphery of the chlorin structure was also sup- and I.I. (21247010), and by a grant from the Ritsumeikan Global ported from the 1H NMR spectral analysis (Supplementary Fig. S2). Innovation Research Organization awarded to Y.K. and H.T. Y. Kashiyama et al. / FEBS Letters 587 (2013) 2578–2583 2583

Appendix A. Supplementary data quantitative signiﬁcance of detoxiﬁcation catabolism of chlorophyll associated with protistan herbivory. Proc. Natl. Acad. Sci. USA 109, 17328–17335. [3] Goericke, R., Strom, S.L. and Bell, M.A. (2000) Distribution and sources of cyclic Supplementary data associated with this article can be found, in pheophorbides in the marine environment. Limnol. Oceanogr. 45, 200–211. the online version, at http://dx.doi.org/10.1016/j.febslet.2013. [4] Engel, N., Curty, C. and Gossauer, A. (1996) Chlorophyll catabolism in Chlorella 06.036. protothecoides. 8 Facts and artifacts.. Plant Physiol. Biochem. 34, 77–83. [5] Kasai, F., Kawachi, M., Erata, M., Mori, F., Yumoto, K., Sato, M. and Ishimoto, M. (2009) NIES-collection list of strains. Jpn. J. Phycol. (Sôrui) 57 (Suppl. 1), 1–350. References eighth ed.. [6] Ma, L. and Dolphin, D. (1999) The metabolites of dietary chlorophylls. [1] Hörtensteiner, S. and Kräutler, B. (2011) Chlorophyll breakdown in higher Phytochemistry 50, 195–202. [7] Falk, H., Hoornaert, G., Isenring, H.P. and Eschenmoser, A. (1975) Enol plants. Biochim. Biophys. Acta 1807, 977–988. 2 3 [2] Kashiyama, Y., Yokoyama, A., Kinoshita, Y., Shoji, S., Miyashita, H., Shiratori, T., derivatives in the chlorophyll series. Preparation of 13 ,17 - Suga, H., Ishikawa, K., Ishikawa, A., Inouye, I., Ishida, K., Fujinuma, D., Aoki, K., cyclopheophorbide enols. Helv. Chim. Acta 58, 2347–2357. Kobayashi, M., Nomoto, S., Mizoguchi, T. and Tamiaki, H. (2012) Ubiquity and