Fe- but not Mg-protophorphyrin IX binds to a transmembrane b-type cytochrome

Abstract

Transmembrane b-type cytochromes, which are crucially involved in electron transfer chains, bind one or more heme (Fe-protoporphyrin IX) molecules non-covalently. Similarly, chlorophylls are typically also non-covalently bound by several membrane integral polypeptides involved in photosynthesis. While both, chlorophyll and heme, are tetrapyrrole macrocycles, they have different substituents at the tetrapyrrole ring moiety. Furthermore, the central metal ion is Mg²⁺ in chlorophyll and Fe^2+/3+ in heme. As heme and chlorophyll a have similar structures and might both be ligated by two histidine residues of a polypeptide chain, and as the local concentration of chlorophyll a might be up to 100-times higher than the concentration of heme, the question arises, as to how an organism ensures specific binding of heme, but not of chlorophyll, to transmembrane apo-cytochromes involved in photosynthetic electron transfer reactions. As shown here, Fe-protoporphyrin IX derivatives with modified substituents at the tetrapyrrole ring moiety still bind to an apo-cytochrome; however, association appears to be reduced. This indicates that hydrophobic and polar interactions of the ring substituents with the protein moiety stabilize the protein/heme-complex but are not essential per se. However, removal or replacement of the central Fe-ion completely abolishes formation of a holo-protein complex, and thus the central iron ion appears to determine heme binding to apo-cytochrome b₆.

• Assembly
• heme
• membrane protein

Introduction

In photosynthetic organelles and organisms, such as chloroplasts and cyanobacteria, several membrane integral proteins are involved in charge transfer reactions across membranes, and many of these proteins contain chlorophyll or heme cofactors (Cramer and Whitmarsh 1977, Gray and Winkler 1996, Schaefer et al. 1996).

The cytochrome b₆f complex is one of the three integral membrane protein complexes of the photosynthetic electron transport chain (Bernát and Rögner 2011, Cramer et al. 2011, Hasan et al. 2013). Heme binding is important for the stability of this complex, since in the absence of heme, the polypeptides of two of the major components of the complex, cytochrome b₆ and cytochrome f, are sensitive to proteolytical degradation (Kuras et al. 1995, 1997, Malnoe et al. 2011). Based on the cytochrome b₆f crystal structures, binding of the two b-type hemes ofcytochrome b₆ via His ligation promotes formation of a closely packed four helix bundle (Kurisu et al. 2003, Stroebel et al. 2003, Hasan et al. 2013), whereas in case of cytochrome f, the free α-amino group of Tyr1 serves as the sixth heme iron ligand and permits close packing of cytochrome f (Martinez et al. 1994). Furthermore, the time point of heme insertion plays an important role during the assembly of the cytochrome b₆f complex dimer and for stability of individual subunits, as reviewed in detail recently (Hasan et al. 2013). Covalent attachment of hemes to c-type cytochromes has been analyzed in detail and four protein systems, catalyzing covalent attachment of hemes to c-type cytochromes, have been identified in pro- and eukaryotes (Thöny-Meyer 1997, de Vitry 2011). In contrast, in case of b-type cytochrome assembly, individual steps involved in heme delivery and/or hemebinding are not understood (Thöny-Meyer 1997). Although sequential binding of hemes to apo-cytochromes has been analyzed in some detail in vitro (as further discussed below), a comprehensive view of principles guiding heme binding to b-type cytochromes has not emerged. As in vitro and in vivo studies indicate spontaneous binding of tetrapyrrole molecules to natural or designed transmembrane (TM) b-type cytochromes (Francke et al. 1999, Cordova et al. 2007, Dreher et al. 2007, Korendovych et al. 2010, Lujan et al. 2012), and as diverse tetrapyrrole macro-cycles might be present at a time in chloroplasts, the question arises how specificity in heme binding is maintained.

In TM b-type cytochromes, which are, e.g. crucially involved in photosynthetic and respiratory electron transfer chains, the heme cofactor is typically ligated by two His residues of the polypeptide chain, which serve as the fifth and sixth ligands of the central iron atom, respectively (von Jagow and Sebald 1980, Yun et al. 1991). However, other residues of the polypeptide chain are also involved in formation of a heme binding pocket and favorable interactions between the heme and the protein drive protein assembly and stabilize the holo-cytochrome (Koch and Schneider 2007). Especially, interactions of the negatively charged heme propionyl groups with the positively charged side-chains of Arg or Lys residues appear to be crucial for heme binding (Cutler et al. 1989, Cross et al. 1995, Prodöhl et al. 2005), as these residues provide positive charges in close proximity to the negatively charged heme propionyl groups (Stroebel et al. 2003, Cramer et al. 2005). Just as hemes in b-type cytochromes, chlorophylls are typically also non-covalently bound by membrane integral polypeptides and the holo-proteins are typically involved in photosynthetic light harvesting and energy transfer. Although more diverse, chlorophyll molecules might also be axially ligated by His residues, as e.g. observed in photosystem II, similar to the axial ligation of hemes in TM b-type cytochromes (Zouni et al. 2001, Kamiya and Shen 2003). In vitro studies have indicated that, similar to TM b-type cytochromes, also chlorophyll containing TM proteins can assemble spontaneously, without the help of additional factors (Paulsen et al. 1990, Cammarata and Schmidt 1992, Horn et al. 2007).

Both, chlorophyll and heme molecules are tetra-pyrrole macrocycles, and both share a common bio-synthetic pathway. However, the two tetrapyrroles have different substituents at the tetrapyrrole ring moiety and, in contrast to heme, chlorophyll a has an additional fifth ring (E-ring) (Willows 2003, Tanaka and Tanaka 2007). Furthermore, the central metal ion is Mg²⁺ in chlorophyll and Fe^2+/3+ in heme. In plants, the biosynthesis pathway of all tetrapyrroles is localized in chloroplasts and only the last steps of heme synthesis are localized in both, mitochondria and chloroplasts (Tanaka and Tanaka 2007). Thus, after their synthesis, both, hemes and chlorophylls, are present in thylakoids and might be incorporated into an appropriate apo-protein. However, as heme and chlorophyll a have similar structures and as the concentration of chlorophyll a might be up to 100-times higher than the concentration of heme (Moulin and Smith 2005), the question arises, as to how an organism ensures specific binding of heme, but not of chlorophyll or chlorophyll precursors, to a TM apo-cytochrome.

In the past few years, progress has been made towards understanding the influence of the heme cofactor on folding and assembly of TM b-type cytochromes. Details of TM b-type cytochrome assembly have been elucidated mainly in vitro by using cytochrome b₆ and the cytochrome b₅₅₉-like protein, cytochrome , as models (Prodöhl et al. 2005, 2007, Dreher et al. 2008). Cytochrome is a homo-dimer of the 44 amino-acid long β subunit (PsbF) of a cyanobacterial cytochrome b₅₅₉ protein (Shinopoulos and Brudvig 2012), which spans a membrane with only a single TM helix. This TM homo-dimer binds one heme cofactor non-covalently, resulting in formation of the holo-cytochrome (Francke et al. 1999, Prodöhl et al. 2005). In the case of the PsbF homo-dimer, integration of the individual TM helices into the membrane and their subsequent interaction is independent of and precedes heme binding (Prodöhl et al. 2005, Volkmer et al. 2006, Akdogan et al. 2012, Weber and Schneider 2013). Additionally, heme binding to a TM b-type cytochrome has been studied using the TM cytochrome b₆ protein, a core subunit of the cytochrome b₆f complex. This four-helix bundle protein binds two heme cofactors non-covalently via four His residues conserved in the TM helices B and D (de Vitry et al. 2004, Hasan et al. 2013). In vivo, a third heme is covalently attached at the Q_i site of apo-cytochrome b₆ via a single cysteine residue (Kurisu et al. 2003, Stroebel et al. 2003, Cramer et al. 2005), but this heme is not of general importance for holo-cytochrome assembly and stability in vivo and in vitro (Dreher et al. 2008, Saint-Marcoux et al. 2009). Apo-cytochrome b₆ integrates into a membrane independent of the cofactor and thereafter appears to adopt a conformation that allows subsequent heme binding in multiple steps (Kuras et al. 1997, Dreher et al. 2007, Dreher et al. 2008, Saint-Marcoux et al. 2009). Mutations of the heme-ligating His residues conserved in helices B and D have indicated that binding of the low-potential heme b_L is a prerequisite for binding of the second, high-potential heme b_H (Kuras et al. 1997, Dreher et al. 2008). Noteworthy, a similar assembly pathway has been suggested before for the cytochrome b subunit of Rhodobacter sphaeroides (Yun et al. 1991).

Based on these findings, a cytochrome b₆ assembly has been proposed, in which at first heme b_L binds in between His86 and His187 located on helix B and D, respectively, whereby His86 appears to have a predominate function. Only after binding of heme b_L the second heme is bound by His100 and His202 (Yun et al. 1991, Kuras et al. 1997, Dreher et al. 2008). Although some details of heme binding to apo-cytochrome b₆ are described, it still is an open question how binding of specifically heme is achieved, whereas unspecific binding of structurally similar tetrapyrroles, such as chlorophyll a, is avoided.

To address this question, we have tested binding of heme, chlorophyll as well as several protoporphyrin IX derivatives, varying in their central ion, to purified apo-cytochrome b₆. While apo-cytochrome b₆ binds heme with high affinity, chlorophyll a is not bound. As heme contains a central Fe-ion, whereas chlorophyll contains Mg²⁺, we tested a potential impact of the central ion on holo-protein formation. In contrast to Fe-pro-toporphyrin IX (heme), the protoporphyrin IX moiety alone (without a central metal ion) as well as Mg-protoporphyrin IX does not bind to apo-cytochrome b₆, and thus the central Fe-ion appears to determine heme binding to apo-cytochrome b₆. Furthermore, Fe-protoporphyrin IX derivatives with modified substituents at the tetrapyrrole ring moiety still bind to the apo-protein. However, association appears to be reduced, indicating that hydrophobic and polar interactions of the ring substituents with the apo-protein stabilize the protein/heme-complex. Interactions of the two heme propionyl groups with the apo-cytochrome appear to be of special importance.

Materials and methods

Expression and purification of cytochrome b₆

Construction of the plasmid pRHMb₆, which codes for the spinach apo-cytochrome b₆ protein N-terminally fused to the maltose-binding protein (MBP) of Escherichia coli, is described in detail elsewhere (Prodöhl et al. 2007). For protein expression, the plasmid pRHMb₆ was transformed into E. coli BL21 (DE3) cells. Het-erologous expression was performed as described in detail recently (Prodöhl et al. 2007). After induction of protein expression, cells were harvested after 3 h and broken by sonication. Subsequently, inclusion bodies were isolated and solubilized in buffer containing 50 mM SDS, prior to protein purification via Ni-NTA affinity chromatography (Prodöhl et al. 2007). The protein was eluted from the Ni-NTA column by decreasing the buffer pH to 4.5. Folding of the protein was initiated by exchanging SDS to the milder, non-ionic detergent n-dodecyl-β-D-maltoside (DDM) (Prodöhl et al. 2007).

Reconstitution of the holo-protein

Heme solutions were prepared exactly as described in Prodöhl et al. (2007). The protoporphyrin IX derivatives Fe(III) protoporphyrin IX dimethylester chloride, Fe(III) deuteroporphyrin IX chloride, Mn (III) protoporphyrin IX chloride, Mg(II) protopor-phyrin IX disodium salt, protoporphyrin IX disodium salt (all from Frontier Scientific, Logan, Utah, USA), Zn(II) Protoporphyrin IX and chlorophyll a (Sigma-Aldrich) were dissolved in 50 mM Tris pH 8, 50 mM NaCl and 5 mM DDM to a final concentration of 0.5 mM. For analyzing binding of the respective porphyrins, these were mixed with apo-cytochrome b₆ in a protein:porphyrin ratio of 1:2. Visible absor-bance spectra were recorded from 350–700 nm with a spectral bandwidth of 0.5 nm on a Perkin Elmer Lambda 35 instrument. Spectra were either taken under air-oxidizing conditions or porphyrins were reduced by addition of sodium dithionite to a final concentration of 5 mM. To quantitatively determine binding of porphyrins to apo-cytochrome b₆, increasing amounts of apo-cytochrome b₆ were titrated into a solution containing 5 μM of the respective porphyrin.

Results and discussion

Apo-cytochrome b₆ binds heme but not chlorophyll a

Although heme and chlorophyll a share a common biosynthesis pathway and have similar structures, their respective affinities to bind to cytochrome b₆ differ significantly. In Figure 1(A), purification of the heterologously expressed cytochrome b₆ protein is shown, and in Figures 1(B), (C), absorbance spectra of in vitro reconstituted apo-cytochrome b₆ are shown under both oxidizing and reducing conditions. Compared to the spectrum of free heme, addition of apo-cytochrome b₆ leads to a clear red shift of 9 and 10 nm of the Soret band under oxidizing and reducing conditions, respectively, and the band width at half maximum is drastically reduced (Table 1). Furthermore, in the redox difference spectrum (inset in Figure 1C) a maximum at 562 nm is pronounced, which is absent in case of free heme. The maxima of the absorbance and redox difference spectra obtained with cytochrome b₆ are prototypical and demonstrate heme binding to the apo-cytochrome and holo-cytochrome b₆ formation, as previously discussed (Prodöhl et al. 2007). While this and previous analyses clearly indicate spontaneous binding of heme to apo-cytochrome b₆, the spectral characteristics of chlorophyll a were not altered at all after addition of apo-cytochrome b₆ (Figure 2, Table 1). Since the spectroscopic properties of chlorophyll a were not affected, and thus the chlorophyll environment was not altered, chlorophyll a does not appear to bind to apo-cytochrome b₆.

Figure 1. Purification and in vitro reconstitution of holo-cytochrome b₆. (A) SDS-PAGE analysis (10% SDS-gel) of the apo-cytochrome b₆ expression and purification. 1: Cell extract before induction; 2: total cellular extract; 3: soluble proteins; 4: inclusion bodies; 5: affinity purified apo-cytochrome b₆ fused to MBP. (B, C) Absorbance spectra show specific binding of heme to apo-cytochrome b₆ under oxidizing (B) and reducing (C) conditions. Vis spectra of free heme (dark grey) and 150 μg in vitro reconstituted apo-cytochrome b₆ (grey) are shown. The inset in Figure (C) shows the redox difference spectrum.

Figure 2. Chlorophyll a does not associate with apo-cytochrome b₆. Absorbance spectra of free chlorophyll a (dark grey) and chlorophyll a after addition of 150 μg cytochrome b₆ (grey) recorded under oxidizing conditions.

Table 1. Absorbance maxima offree protoporphyrin IX derivatives (without cyt b₆) and in presence ofcytochrome b₆ (with cyt b₆) are listed. The bandwidth at half maximum (fwhm) is given for the Soret band (γ-band) of each spectrum. Absorbance maxima and the fwhm values were obtained after Gaussian deconvolution of the respective absorbance curve using the program PeakFit (Systat Software, Inc.).

		Without cyt b6		With cyt b6
Porphyrin	Redoxstate	Maximum	Fwhm	Maximum	Fwhm
Fe-protoporphyrinI X (heme)	Oxidized	405	32.7	414	26
		606		531
				562
	Reduced	418	39.8	429	22
		541		532
		569		562
Chlorophyll a	Oxidized	378		378
		416		416
		439	24.7	439	23.4
		588		590
		626		627
		671	25.7	671	24.5
Protoporphyrin IX	Oxidized	408	38.2	408	39
		506		506
		541		541
		578		578
		632		632
Mg-protoporphyrin IX	Oxidized	419	23.7	419	23.3
		551		551
		588		588
Fe-deuteroporphyrin IX	Oxidized	401	33.5	409	33
		569		570
		608		605
	Reduced	408	34.3	411	33.8
		520		525
		554		549
Fe-protoporphyrin IX dimethylester	Oxidized	395	22.5	400	23
		488		524
		592
	Reduced	409	39.9	425	21.7
		569		530
		606		559

To further test binding of heme vs. chlorophyll a to apo-cytochrome b₆, we purified the protein a second time after addition of the two different cofactors and monitored whether the cofactors co-eluted with the protein (Figure 3). Solely upon addition of heme, a spectrum of a cofactor containing holo-cytochrome was monitored, whereas in the case of chlorophyll the cofactor did not co-elute with the protein, which further demonstrates that chlorophyll a does not bind to the apo-protein. However, this observation begs the question, which structural features of Fe-protoporphyrin IX determine its selective binding to apo-cytochrome b₆.

Figure 3. Heme but not chlorophyll co-elutes with cytochrome b₆ after in vitro reconstitution. Free heme and chlorophyll a, respectively, were added to isolated apo-cytochrome b₆ and the protein was purified via Ni-NTA chromatography. After elution of the protein, absorbance spectra were recorded under oxidizing (dark grey) and reducing conditions (grey) in case of heme. Chlorophyll a did not elute from the column together with the protein, therefore a chlorophyll a spectrum is not visible (light grey).

The central iron ion is essential for porphyrin binding to apo-cytochrome b₆

In order to identify structural determinants driving specific binding of heme, but not of chlorophyll a to apo-cytochrome b₆, binding of commercially available protoporphyrin IX derivatives, having different central metal ions, to apo-cytochrome b₆ was analyzed (Figure 4 and Supplementary Figure S1, available online). The absorbance spectrum of free protopor-phyrin IX (without a central metal ion) has characteristic maxima at 409, 506, 541, 578 and 632 nm. Addition of apo-cytochrome b₆ to protoporphyrin IX did neither influence the maxima of protoporphyrin IX absorbance spectrum nor the band widths at half maximum (Table 1). Thus, the protoporphyrin IX environment is not altered (Figure 4), and hence apo-cytochrome b₆ does not bind protoporphyrin IX. As the heme central ion interacts with the heme-ligating His residues of cytochrome b₆, this observation assorts well with the recent observation that heme binding to apo-cytochrome b₆ can be disturbed when the heme ligating His residues are mutated to Ala (Kuras et al. 1997, Dreher et al. 2008). Thus, these data indicate that binding of a metal center by the cytochrome b₆ His residues is crucial for porphyrin binding to apo-cytochrome b₆. To subsequently monitor interactions of the protein with a metal center in greater detail, we next analyzed binding of the chlorophyll synthesis intermediate Mg-protoporphyrin IX, which differs from heme only in its metal center.

Figure 4. Apo-cytochrome b₆ does not bind porphyrins with central and porphyrins in presence of 145 μg apo-cytochrome b₆ (grey) and porphyrins in presence of 145 mg apo-cytochrome b₆ (grey) are show nunder oxidizing conditions. (A) ProtoporphyrinIX, (B) Mg-protoporphyrin IX.

As shown in Figure 4, the absorbance maxima and the band width at half maximum of the free porphyrin were not influenced by addition of apo-cytochrome b₆ (Table 1), and thus replacement of Fe^2+/3+ by Mg²⁺ already eliminates binding of the protoporphyrin IX molecule. In addition, we have also analyzed binding of protoporphyrin IX derivatives, which have zinc or manganese as central metal ions (Supplementary Figure S1) and also these analyzed protoporphyrins do not appear to interact with the apo-cytochrome, which further highlights the special role of the Fe^2+/3+-center. While the central Fe^2+/3+ ions of cytochromes are typically hexa-coordinated in vivo, the Mg²⁺ ions of most chlorophylls are only penta-coordinated (Cogdell et al. 2003, Smith et al. 2010). Because of this, the position of the Mg²⁺ ion is slightly out of the equatorial plane, and it has been argued that a sixth axial ligand would be energetically less favorable (Storm 1970). Thus, the preferred penta- vs. hexa-coordination state might explain the observed differences in chlorophyll vs. heme binding described here. However, as the Mg²⁺ ions of a certain fraction of chlorophylls are also hexa-coordinated in vivo, other factors are most likely as well crucial for binding of defined tetrapyrroles in vivo. These factors might include, but are not limited to, the His ligand side-chain rotamers (Braun et al. 2011) and the different tetrapyrrole ring substituents (as further addressed below).

Taken together, these data indicate that specifically Fe-protoporphyrin IX interacts with the apo-cyto-chrome, and replacement by a different metal ion or its complete removal prevents distinct binding to apo-cytochrome b₆. This leads to the assumption that it is the central iron ion, which is crucial for specific binding of Fe-protoporphyrin IX to apo-cytochrome b₆, which already nicely elucidates why cytochrome b₆ specifically binds Fe-protoporphyrin IX in vivo but not chlorophyll or chlorophyll precursors, as further discussed below.

The propionyl and ethylene side chains of protoporphyrin IX affect binding of Fe-protoporphyrin IX to apo-cytochrome b₆

While the above presented results suggest that the central metal ion is crucial for binding of Fe-protoporphyrin IX to the TM apo-cytochrome, a potential impact of the tetrapyrrole side chains on the stability of cytochrome b₆ remains open. Based on previous analyses, apo-cytochrome b₆ folds into a structure, which allows binding of the heme cofactors, resulting in structural rearrangements and formation of the holo-cytochrome (Dreher et al. 2007, 2008, Prodöhl et al. 2007). Many forces guide formation and stability of the holo-cytochrome, and the entropy loss, which occurs when the protein interacts with the cofactor, is minimized by formation of favorable interactions between the protoporphyrin cofactor and the protein, resulting in an enthalpy gain and finally in formation of the holo-cytochrome. Such favorable interactions might include hydrophobic and van der Waals interactions, as well as formation of hydrogen bonds or salt bridges.

To identify the impact of individual tetrapyrrole ring substituents on binding to the apo-cytochrome, we analyzed binding of two Fe-protoporphyrin IX derivatives (Figures 5 and 6). In case of deuteropor-phyrin IX, the two ethylene groups at rings A and B, which might establish hydrophobic contacts and stabilize van der Waals interactions, are replaced by hydrogens. When apo-cytochrome b₆ was added to a solution containing Fe-deuteroporphyrin IX, differences in the absorbance spectra were observed (Figure 6A, 6B). Under oxidizing conditions the g-maximum of Fe-deuteroporphyrin IX in presence of apo-cytochrome b₆ is clearly red-shifted by 5 nm compared to the free porphyrin, and the peak intensity increases in presence of the protein. Furthermore, the pronounced α-band maximum of free Fe-deuteroporphyrin IX at 595 nm disappeared after addition of the apo-cytochrome b₆. Under reducing conditions, a 4 nm red shift of the g-maximum was observed after addition of the apo-cytochrome b₆ together with a rise in the absorbance intensity, when compared to the free porphyrin. Thus, the observed spectral changes (summarized in Table 1) clearly indicate that this porphyrin interacts with the protein moiety. In the redox difference spectra (inset in Figure 6B), free Fe-deuteroporphyrin IX has a broad absorbance maximum between 500 and 580 nm without any clearly defined maxima. In contrast, after addition of apo-cytochrome b₆ to Fe-deuteroporphyrin IX, a maximum at 547 nm appeared, demonstrating specific binding of Fe-deuteroporphyrin IX to apo-cytochrome b₆ and holo-cytochrome formation. These results lead to the conclusion that the ethylene substituents at the tetrapyrrole ring moiety are not essential per se for heme binding. However, compared to the spectra obtained after in vitro reconstitution of cytochrome b₆ with heme (Figure 1B), the spectra were less well defined and the width at half maximum of the Soret band was not reduced, which might indicate destabilization of the holo-cytochrome due to reduced cofactor binding, as further analyzed below.

Figure 5. Chemical structures ofheme, chlorophyll a and protoporphyrin IX derivatives used in this study. Arrows indicate side chains varied compared to heme. In Fe-deuteroporphyrin IX ethylene side chains 2 and 4 are substituted for hydrogen atoms whereas in Fe-protoporphyrin IX dimethylester the propionyl groups 6 and 7 are esterified with methyl groups.

Figure 6. Apo-cytochrome b₆ binds porphyrins with modified side chains. Absorbance spectra of Fe-deuteroporphyrin IX (A, B) as well as of Fe-protoporphyrin IX dimethylester (C, D) were recorded under oxidizing (A, C) and reducing (B, D) conditions. The spectrum of free porphyrin (dark grey) is compared to the spectrum of the porphyrin after addition of 150 μg apo-cytochrome b₆ (grey). The insets in (B) and (D) are redox difference spectra.

Interaction of the heme propionyl side chains with positively charged amino acid side chains of the protein are discussed to be of special importance for heme binding, and thus for assembly of the holo-cytochrome. Therefore, the second analyzed protoporphyrin IX derivative, the Fe-protoporphyrin IX dimethylester, possesses esterified propionyl side chains at rings C and D (Figures 5 and 6). Under oxidizing conditions, the γ-maximum of the free porphyrin is shifted by 8 nm after the addition of apo-cytochrome b₆ (Table 1). Under reducing conditions, the free porphyrin shows a broad γ-band at around 400 nm and β- and α-band maximum at 569 nm and 601 nm, respectively. The α-maximum disappeared upon addition of apo-cytochrome b₆, and a well-defined γ-maximum at 425 nm, a β-band at 530 nm and an α-band maximum at 559 nm was pronounced. Under both, oxidizing and reducing conditions the absorbance intensity of the γ-maximum rose after addition of the apo-cytochrome b₆, which was however more obvious under reducing conditions. Thus, after addition of apo-cytochrome b₆, a well defined spectrum was observed under reducing conditions, which is rather different from the spectrum of free Fe-protoporphyrin IX dimethylester. Furthermore, the redox difference spectrum shows two distinct maxima at 530 nm and 560 nm (inset, Figure 6D), in contrast to the free porphyrin. These differences between the spectral properties demonstrate a specific interaction between Fe-protoporphyrin IX dimethylester and apo-cytochrome b₆. However, binding of Fe-protoporphyrin IX dimethylester seems to be weaker compared to Fe-protoporphyrin IX, and the band widths at half maximum do not change that significantly (Table 1).

Therefore, to identify differences in binding of these two heme derivatives to the apo-protein in greater detail, we performed titration experiments and measured porphyrin binding to apo-cytochrome b₆ by absorbance spectroscopy at different porphyrin-to-protein ratios (Figure 7). At a ratio of about 2 hemes per protein (i.e. a protein-to-heme ratio of 0.5) the absorbance does not change drastically anymore, indicating that one apo-cytochrome b₆ protein binds 2 heme molecules with high affinity, as already analyzed in detail previously (Prodöhl et al. 2007, Dreher et al. 2008). Furthermore, the titrations also show that apo-cytochrome b₆ associates more weakly with the heme derivatives Fe-deuteroporphyrin IX and Fe-protoporphyrin IX dimethylester. In addition, the sigmoidal curve observed in case of protoporphy-rin IX dimethylester suggests that binding of (at least) protoporphyrin IX dimethylester to apo-cytochrome b₆ is highly cooperative, i.e. binding of the first heme significantly promotes binding of the second heme. These findings are supported by previous mutational analyses, which have revealed that binding of the low potential heme b_L has to occur prior to binding of the high potential heme b_H, and thus binding of the first heme is a prerequisite for binding of the second heme (Kuras et al. 1997, Dreher et al. 2008).

Figure 7. Porphyrin binding isotherms for cytochrome b₆. Increasing amounts ofapo-cytochrome b₆ were titrated into buffer containing either 5 μM heme (squares), Fe-deuteroporphyrin IX (circles) or Fe-protoporphyrin IX dimethylester (triangles). The protein-to-heme ratio was plotted against the normalized absorbance change at 413 nm. Each data point represents an individual measurement.

Together, the results indicate that interactions of the heme propionyl group with positively charged residues of the polypeptide chain are crucial for high affinity binding of the heme cofactor. Noteworthy, while ionic interactions between the positively charged amino acid side chain and the negatively charged heme propionyl groups are abolished in the case of the analyzed ester, formation of stabilizing hydrogen bonds to the ester carbonyl oxygen is still possible, which might ensure binding of the dimethylester derivative. Furthermore, replacement of the two ethylene groups affects the stability of the protein:heme complex, most likely since favorable van der Waals interactions between the two heme ethylene groups and the protein moiety are eliminated, resulting in less enthalpy gain after binding of the protoporphyrin.

Conclusions

In plants, all tetrapyrroles share a common biosynthesis pathway, which is localized in plastids, and only the last two steps of heme synthesis are localized in mitochondria as well (Cornah et al. 2003). The first precursor, 5-aminolaevulinic acid, is generated from glutamate. Two molecules are condensed to porpho-bilinogen, which thereafter polymerizes to form the linear tetrapyrrole 1-hydroxymethylbillane. Ring closure is followed by several decarboxylation and oxidation reactions at the tetrapyrrole macrocycle, finally leading to the formation of protoporphyrin IX (Beale 1999, Tanaka and Tanaka 2007). At this point chlorophyll and heme synthesis pathways split up. Insertion of Mg²⁺ by the Mg-chelatase generates Mg-protoporphyrin IX, which is further converted in multiple steps to chlorophyll. Heme is directly synthesized from protoporphyrin IX by insertion of Fe²⁺ catalyzed by the ferrochelatase. The data presented here indicate that the substituents at the tetrapyrrole ring influence association with a TM apo-cyto-chrome. Typically, heme and chlorophyll cofactors bind at pre-defined binding cavities to the protein moiety. Thus, as chlorophyll a contains a long phytyl side chain, this large ring substituent might already abolish binding of chlorophyll to apo-cytochromes in vivo due to steric clashes. However, binding of the chlorophyll synthesis intermediate Mg-protoporphyrin IX would then not be abolished. Based on the presented results, incorporation of por-phyrins into apo-cytochrome b₆ is only possible after the last step of heme synthesis, i.e. after insertion of the central iron ion, is completed. Free protoporphy-rin IX, as well as the chlorophyll synthesis intermediate Mg-protoporphyrin IX, are not able to bind to the cytochrome b₆ heme binding sites. Noteworthy, the heme iron center is also crucial for functional assembly of the E. coli TM cytochrome b₅₅₆ protein (Nihei et al. 2001), and while E. coli does not synthesize chlorophyll, in this case the iron-center appears to be also crucial for binding of a protopor-phyrin molecule. Thus, the Fe-chelatase is the key enzyme in the heme biosynthesis pathway, which in the end determines the fate of the protoporphyrin IX moiety and thereby might control assembly of TM b-type cytochromes.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Acknowledgements

This work was supported by grants from the Ministry of Science, Research and Arts of Baden-Württem-berg, from the Centre of Complex Matter and from the University of Mainz. We thank Daniel Otzen for helpful discussions and Hildegard Pearson for critically reading the manuscript.