INTRODUCTION

Top Abstract Introduction Procedures Results & Discussion References

The phenomenon of chlorophyll (Chl)¹ breakdown in degreening plants has long been a topic of general biological interest (1). Although an estimated 1 billion tons of Chl are degraded annually on earth, the biochemical pathway of Chl catabolism has remained poorly defined because of difficulties in extracting and identifying the catabolic products of the porphyrin moiety (2). Only in the last few years have the colorless end products of catabolism been discovered in senescent leaves of several plant species (3-7). These nonfluorescent Chl catabolites (NCCs) share a 1-oxo-19-formylbilane structure (see Fig. 1). Oxidative cleavage at the C4/C5 meso position of the chlorin macrocycle appears to be a common step in Chl degradation and the formation of NCCs. NCCs formed in different organisms have the same basic tetrapyrrolic skeleton but differ from each other by species-specific modifications of the side chains (3-7).

The key step in Chl breakdown in green plants, the cleavage reaction of the porphinoid macrocycle, is catalyzed by an oxygenase that specifically recognizes pheophorbide a (Pheide a) (8). This iron-containing enzyme, Pheide a oxygenase (PaO), is located in the envelope of senescent chloroplasts (9). The oxygenase is the key enzyme in Chl catabolism; in two cases of stay-green mutants, one of them Mendel's "green peas," the genetic lesion responsible for high Chl retention during leaf senescence was shown to result in reduced PaO activity (10, 11). The conversion of Pheide a to a primary blue fluorescent catabolite (pFCC; Ref. 12) requires the joint action of PaO and a soluble stroma-located enzyme that reduces an intermediary red catabolite (RCC; Ref. 13) to pFCC (Fig. 1, Ref. 14). Both PaO and RCC reductase require reduced ferredoxin (Fd) as reductant. Although RCC reductase was found to be present at all stages of development and in all tissues examined (15), PaO seems to occur in gerontoplasts exclusively (16).

Chl breakdown has also been studied in the green alga Chlorella protothecoides. This alga excretes red pigments from the cells when grown in a glucose-rich, nitrogen-deficient medium (17, 18). These pigments were characterized as a mixture of bilins with the structure of 1-oxo-19-formylbilanes (18, 19), suggesting a similar mechanism of oxygenolytic cleavage (at C4/C5 of the porphinoid macrocycle) in green algae and degreening plants (6, 19). ¹⁸O-Labeling experiments with cultures of C. protothecoides that were degrading Chl demonstrated that only one of the two O atoms incorporated into the red pigments was derived from dioxygen (at the formyl group attached to pyrrole B) (20). These studies strongly suggest that the oxygenolytic cleavage is performed by a monooxygenase (19, 20).

Here we report experiments that examine the basic mechanism of the oxygenolytic cleavage of Pheide a by PaO in the dicot Brassica napus. Our major goal was to determine whether PaO from higher plants is a monooxygenase, as suggested for its algal counterpart. Because ¹⁸O-labeling in vivo was not feasible, an in vitro system employing partially purified PaO from senescent canola cotyledons was used to determine the mechanism of the oxygenolytic cleavage.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results & Discussion References

Cultivation of Plants

Seedlings of canola, B. napus L. cv. Lirajet, were grown as described (8).

Isolation of PaO and RCC Reductase

Chloroplast membranes were prepared as described (8). The supernatant of the first washing step was employed as the source of RCC reductase in the labeling experiments. Washed gerontoplast membranes were diluted to a [Chl] of 3 mg ml ¹, solubilized with 1% (v/v) Triton X-100, and, after centrifugation at 150,000 × g for 1 h (8), PaO was partially purified from the supernatant by chromatography on EAH-activated Sepharose (Amersham Pharmacia Biotech) (12). The supernatant was loaded onto the EAH-activated Sepharose column that had been equilibrated with 25 mM Tris-Mes, pH 8.0, 0.1% Triton X-100, the column was washed with the same buffer containing 0.1 M KCl (10 column volumes), and the PaO activity was eluted with 0.2 M KCl in equilibration buffer. Active fractions were pooled (EAH-Sepharose pool) and concentrated 5-fold by ultrafiltration employing Centriplus-50 concentrators (Amicon). PaO activity was monitored in a standard assay (8), and protein concentrations were determined by the method of Bradford (21) following the protocol of Peterson (22).

¹⁸O-Labeling Experiments

The assay mixtures contained 0.5 ml of partially purified PaO (equivalent to 20 g fresh weight), 0.5 ml RCC reductase (equivalent to 20 g fresh weight), 30 µl each of NADPH (60 mM), glucose-6-phosphate (100 mM), glucose-6-phosphate dehydrogenase (10 units ml ¹), and Fd (10 mg ml ¹). These components were pipetted into a 10-ml flask equipped with a three-way stopcock to allow evacuation and gassing. The flask was placed on ice and purged with N₂ for 1 min followed by the addition of Pheide a to a final concentration of 1 mM. After purging the reaction mixture for an additional 5 min, the flask was evacuated (vacuum pump) for 5 min and than aerated with ¹⁸O oxygen (97%; Cambridge Isotope Laboratories). After incubating the air-tight flask for 2 h at room temperature in the dark, the reaction was stopped by the addition of 1.5 ml of ice-cold methanol. The solution was clarified by centrifugation (12,000 × g; 5 min) and then stored in liquid N₂. Identical incubations were performed either in the presence of air (i.e. ¹⁶O oxygen) or under reduced pressure (i.e. under vacuum).

Purification of Reaction Products

¹⁸O-Labeled pFCC-1 (Bn-FCC-2) was isolated from the pooled reaction mixtures by reverse-phase HPLC (10). After two cycles of purification, pFCC-1 was desalted on a C18-SepPak cartridge (Waters) and lyophilized; the yield was approximately 30 µg of ¹⁸O-pFCC, as determined spectrophotometrically (10). ¹⁶O-pFCC-1 formed under ¹⁶O₂ atmosphere was isolated in an identical manner; the yield was approximately 25 µg. For analysis by FAB-MS (Finnigan MAT-95, cesium gun at 20 keV; 2 µA; positive-ion mode; glycerin matrix), the samples of pFCCs were dissolved in methanol.

¹⁸O-pFCC-1-- m/z (%) = 633.4 (15), 632.1 (44), 631.1 (100, C₃₅H₄₁O₆¹⁸O [M+1]⁺), 630.1 (36), 629.2 (18), 511.3 (13), 510.4 (40), 509.2 (59, [M + 1  C₇H₈NO]⁺ = [M + 1  ring A]⁺), 493.0 (15), 492.2 (34, [M + 1  C₈H₁₁¹⁸O]⁺ = [M + 1  ring B]⁺), 490 (9).

¹⁶O-pFCC-1-- m/z (%) = 631.1 (18), 630.1 (46), 629.1 (100, C₃₅H₄₁N₄O₇ [M + 1]⁺), 628.3 (27), 627.3 (13), 509.1 (15), 508.3 (45), 507.0 (50, [M + 1  C₇H₈NO]⁺ = [M + 1  ring A]⁺), 506.2 (25), 505.2 (11), 493.1 (14) 492.3 (28 [M + 1  C₈H₁₁O]⁺ = [M + 1  ring B]⁺), 491.3 (12).

	RESULTS AND DISCUSSION

Top Abstract Introduction Procedures Results & Discussion References

Chlorophyll breakdown in higher plants proceeds via an oxygenolytic opening of the porphinoid macrocycle that is essentially irreversible (3). This reaction is catalyzed by an oxygenase only detectable during leaf senescence. Present evidence supports the existence of a single major pathway of Chl breakdown in green plants in which Pheide a is converted, via RCC, to fluorescing catabolites, such as pFCC-1. The FCCs are further rapidly modified and deposited in the vacuoles as colorless, nonfluorescing tetrapyrroles (such as Bn-NCC-3, see Fig. 1). The biochemical mechanism of the decisive oxygenolytic cleavage of Pheide a in senescent plants is of particular interest. However, it is not possible to perform specific in vivo ¹⁸O-labeling from molecular oxygen, the suspected source of the oxygen atoms for the oxygenolytic cleavage, because high rates of respiration in senescing leaves would lead to the accumulation of labeled water. As an alternative approach, we performed enzymic cleavage of Pheide a in vitro using partially purified PaO; this would exclude interference with other oxygen-consuming reactions. The experiments were designed to yield isolatable quantities of pFCC-1 rather than RCC because accumulation of the latter inhibits the oxygenase reaction (14). According to a formal analysis, the formation of RCC from Pheide a involves, overall, the addition of two atoms each of oxygen and hydrogen, whereas the conversion from RCC into pFCC-1 requires two additional atoms of hydrogen (12, 13).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Structures of breakdown products of Chl. Two-step cleavage of Pheide a to a primary FCC (pFCC-1) with RCC as an intermediate and further transformation into NCCs is shown. NCCs isolated from higher plants have a uniform basic constitution; known variation of substituents R1, R2, R3 of NCCs from different plant species is indicated (3-7).

To minimize nonspecific oxygen consumption by residual oxygen-dependent reactions likely to be present in the reaction assays employed for ¹⁸O-labeling experiments, PaO was partially purified from senescent canola cotyledons (Table I). Because the enzyme is localized in gerontoplast membranes (8, 9), it was possible to achieve a 6-fold purification by washing of membranes three times, solubilizing the activity with Triton X-100, and sedimenting Chl·protein complexes by centrifugation at 150,000 × g. After a chromatographic separation on EAH-activated Sepharose, PaO activity was enriched 121-fold, whereas 86% of the total initial activity was recovered.

When partially purified PaO was incubated with RCC reductase in an in vitro assay containing Fd and a Fd-reducing system (NADPH, glucose-6-phosphate, glucose-6-phosphate dehydrogenase), about the same quantities of pFCC-1 were formed from Pheide a in the presence of ¹⁶O₂ oxygen and ¹⁸O₂ oxygen, respectively. In contrast, the yield of pFCC-1 was reduced to only 9% when the incubation was performed under vacuum; the low level of pFCC-1 synthesis was probably due to residual oxygen present in the reaction mixture even after flooding the reaction flask with N₂ and subsequent evacuation. ¹⁸O-Labeled and unlabeled pFCC-1 were purified by reverse-phase HPLC with a recovery of approximately 83 and 72%, respectively. As judged by their chromatographic behavior and spectroscopic properties (UV absorption and fluorescence), the samples were identical to authentic pFCC-1 (Bn-FCC-2) synthesized under natural atmospheric conditions (8, 12) (data not shown).

A molecular mass of m/z = 629.301 ± 0.004 was determined by high resolution FAB-MS for the authentic Bn-FCC-2, corresponding to a molecular formula of C₃₅H₄₁N₄O₇ (12). The identical result was obtained upon FAB-MS of ¹⁶O-pFCC-1 with an experimental m/z of 629.1 (Fig. 2 bottom). In the presence of ¹⁸O_2, cleavage of Pheide a yielded a pFCC-1 with m/z = 631.1 (Fig. 2, top) for the molecular ion. The signal pattern of the molecular ion was similar in both cases, and the isotopic composition of ¹⁸O-pFCC-1 for oxygen could be estimated as 5% ¹⁶O₇, 95% ¹⁸O¹⁶O_6, and 0% ¹⁸O₂¹⁶O₅ (3% S.D.). Accordingly, one of the two oxygen atoms introduced into pFCC-1 during ¹⁸O-labeling is derived from the ¹⁸O₂ oxygen gas supplied.

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Section of the FAB-mass spectrum showing the molecular ion of Bn-FCC-2 18O = 18O-pFCC-1 (top) and of Bn-FCC-2 = 16O-pFCC-1 (unlabeled, bottom). For experimental details see text.

In the mass spectra of both ¹⁸O-pFCC-1 and ¹⁶O-pFCC-1, a fragment ion was observed (at m/z = 507.1 versus m/z = 509.1) to arise from the molecular ion by loss of 122 mass units, corresponding to loss of an unlabeled ring A-unit (C₇H₈NO) (Fig. 3). In contrast, in both mass spectra a common fragment was found at m/z = 492.1, indicating loss of the ring B-unit as C₈H₁₁N¹⁶O (137 mass units) and C₈H₁₁N¹⁸O (139 mass units), respectively, in the spectra of ¹⁶O-pFCC-1 and ¹⁸O-pFCC-1. These results show that ¹⁸O incorporation had occurred at the formyl group attached at pyrrole B, whereas the lactam oxygen of ring A is not derived from molecular oxygen.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   FAB-mass spectrum showing mass regions between m/z = 475-645 of Bn-FCC-2 18O = 18O-pFCC-1 (top) and of Bn-FCC-2 = 16O-pFCC-1 (unlabeled, bottom). Note the fragment ion at m/z = 492 (due to loss of ring B) and at m/z = 509 (top) and 507 (bottom), respectively, due to loss of ring A.

The mass spectrometric analysis of ¹⁸O-pFCC-1 revealed the incorporation of one oxygen atom derived from molecular oxygen, indicating that PaO catalyzes the regioselective attachment of that oxygen atom at the C5 position of Pheide a. As to the fate of the second O atom of dioxygen, which is not incorporated into the catabolite, the in vitro system may provide a preliminary answer because it is defined in terms of the cofactors added to the incubation mixture. It has been demonstrated that PaO isolated from senescent leaf tissues is an iron-containing enzyme that requires reduced Fd for the operation of its redox cycle (23). There seems little doubt that the second O atom of dioxygen is reduced to water and that the two electrons required are ultimately derived from reduced Fd (Fig. 4). On the other hand, the second oxygen atom that appears in the catabolite as the lactam oxygen at C4 of ring A is in all probability derived from water.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Pheide a cleavage to a red Chl catabolite as catalyzed by Pheide a (mono)oxygenase. Note that during 18O-labeling only the formyl group at ring B is labeled. The direct source of the lactam oxygen remains unclear, but most probably is H2O.

We conclude that PaO, the enzyme that opens the ring of Pheide a during Chl breakdown in degreening plants is a monooxygenase that delivers an oxygen atom to the C5 position of the Pheide a macrocycle. Earlier ¹⁸O-labeling experiments with the green alga C. protothecoides yielded evidence for the incorporation of a single atom of dioxygen into the tetrapyrrolic cleavage products. The algal enzyme responsible for the ring cleavage reaction has correspondingly been suggested to be a monooxygenase (20). It was further proposed that oxygenation proceeds via formation of an epoxide at position C4/C5, which is subsequently hydrolyzed to yield upon intramolecular rearrangement a red bilin (related to RCC) in which the -methine bridge is saturated (19, 20). This mechanism and the formation of an epoxide intermediate remain to be verified.

It is not known whether the oxygenase reaction in C. protothecoides requires reduced Fd, as is the case in angiosperms. However, the algal enzyme certainly appears to be distinct from its counterpart in higher plants in several respects. First, in the alga, the red tetrapyrrolic cleavage product is released and eventually excreted into the culture medium. In the case of higher plants, the release of the oxygenation product RCC from PaO requires the action of a second enzyme, RCC reductase, which catalyzes the reduction of the -methine bridge to yield pFCC (12). The formation of pFCC is the first in a series of modifications and conjugations that result in the synthesis of NCCs, which are finally deposited in the vacuoles of senescent mesophyll cells (24, 25). Furthermore, the monooxygenases of C. protothecoides and of higher plants differ from each other in their substrate specificity. Although the natural substrate of the cleavage reaction (as well as its product) has yet to be elucidated in the alga, red pigments derived from both Chl a and b have been isolated from the medium of degreened algal cultures (26). In contrast, all catabolites that have been isolated from higher plants have been shown to be derived from Chl a (3-7) and, in addition, the porphyrin-cleaving monooxygenase from canola specifically recognizes Pheide a but not b as substrate (8).

These differences between angiosperms and algae not withstanding, our findings support the notion of a common basic cleavage path for Chl in the plant kingdom. It seems likely that the type of monooxygenase responsible for the oxygenation of Pheide a in a higher plant such as B. napus is similar to that of the phylogenetically much older green alga. We speculate that this mode of Chl breakdown evolved at an early stage in the evolution of autotrophic organisms and is perhaps (nearly) as old as photosynthesis itself. One major difference between the algae and vascular plants is the way they eliminate chlorophyll catabolites. In the alga, the catabolites are excreted, whereas vascular plants have developed a more complex mechanism of intracellular detoxification. The coupling of Pheide a monooxygenase to RCC reductase represents the initial step in this surprisingly complicated process, whereby Chl is ultimately sequestered in the form of colorless catabolites, which are deposited in the vacuole.