On the mode of integration of the thylakoid membrane protein cytochrome b 6 into cytoplasmic membrane of Escherichia coli

In the stroma compartment, several pathways are used for integration/translocation of chloroplast proteins into or across the thylakoid membrane. In this study we investigated the mode of incorporation of the chloroplastencoded cytochrome b6 into the bacterial membrane. Cytochrome b6 naturally comprises of four transmembrane helices (A,B,C,D) and contains two b-type hemes. In the present study, mature cytochrome b6 or constructed deletion mutants of cytochrome were expressed in E. coli cells. The membrane insertion of cytochrome b6 in this bacterial model system requires an artificially added presequence that directs the protein to use an E. coli membrane-insertion pathway. This could be accomplished by fusion to maltose-binding protein (MBP) or to the bacterial Sec-dependent signal peptide (SSpelB). The integration of mature cytochrome b6 into the bacterial cytoplasmic membrane by the Sec pathway has been reported previously by our group (Kroliczewski et al., 2005, Biochemistry, 44: 7570). The results presented here show that cytochrome b6 devoid of the first helix A can be inserted into the membrane, as can the entire ABCD. On the other hand, the construct devoid of helices A and B is translocated through the membrane into the periplasm without any effective insertion. This suggests the importance of the membrane-anchoring sequences that are likely to be present in only the A and B part, and it is consistent with the results of computational prediction which did not identify any membrane-anchoring sequences for the C or D helices. We also show that the incorporation of hemes into the truncated form of cytochrome b6 is possible, as long as the B and D helices bearing axial ligands to heme are present.


INTRODUCTION
Membrane proteins account for 20-25 % of all open reading frames in sequenced genomes, and fulfill a wide range of central functions in cells (Wallim & von Heijne, 1998).Our knowledge regarding the assembly of this very important class of proteins is still poor and the functional and structural studies of membrane proteins have been strongly hampered due to the difficulties in overexpressing them in high yields.Little is also known about the mechanisms of targeting and insertion of chloroplast-encoded thylakoid membrane proteins despite their importance in chloroplast function (Rochaix, 1996;Choquet et al., 1998).Chloroplasts are the site of photosynthesis and thylakoids are responsible for the crucial processes of light harvesting, electron transport and photophosphorylation.It is worth emphasizing that the chloroplast thylakoid is one of the most abundant membrane systems in nature.The structural and functional complexity of the chloroplast indicates that chloroplast proteins must be transported exactly to the correct compartment to reach their functional locations.Because chloroplasts are believed to have descended from prokaryotes, it can be postulated that the targeting and insertion pathways of the chloroplast-encoded proteins are likely to show strong similarities to those in the Gram-negative bacterium Escherichia coli (Schatz & Dobberstein, 1996).Indeed, several chloroplast homologues of E. coli targeting and insertion components have been identified (Schunemann, 2007;Cline & Dabney-Smitch, 2008;Aldridge et al., 2009).The insertion of some thylakoid membrane proteins (cytochrome f, cytochrome b 6 ) expressed in E. coli into the bacterial cytoplasmic membrane has also been demonstrated (Rothstein et al., 1985, Kroliczewski et al., 2005).
Commonly, proteins are targeted to the membrane by hydrophobic amino-acid sequences which are either transmembrane segments (TMs) of membrane proteins or cleavable signal sequences (SSs).Typically, receptors located within the membrane recognize the hydrophobic amino-acid sequence and the membrane-embedded translocation machinery promotes the insertion into or transport across the membrane.
It is thought that in the chloroplast stroma, four transport pathways are used for the targeting of chloroplast proteins into or across the thylakoid membrane (Robinson et al., 2001).These pathways are named the cpSec-dependent, cpSRP-dependent, DpH-dependent, and spontaneous (unassisted) pathways (Cline et al., 1992;Sundberg et al., 1997;Woolhead et al., 2001;Yi & Dalbey, 2005;Aldridge et al., 2009), and all of them have direct cognates in bacteria (Schatz & Dobberstein, 1996).The chloroplast cpSec pathway evolved from the secretory pathway involved in export of Sec-dependent proteins to the periplasm in bacteria.In E. coli the functional translocon of the Sec-dependent pathway contains at least three components: SecA, SecY and SecE.In post-translational transport, the SS of the preprotein interacts with SecA in the cytoplasm and the formed complex (SecA-preprotein) associates with the membrane components composed of SecY and SecE integral proteins which form the transmembrane channel (Aldridge et al., 2009).The channel itself is a passive conduit for transported polypeptides and SecA is an ATPase that provides the driving force for translocation of the protein through the Sec channel pore.SecA is also a binding partner of SecY.Generally, it is accepted that the SecY core channel complex is a heterotrimer consisting of SecY, SecE and SecG subunits (Rapoport et al., 1996), which was confirmed by a crystal structure study of detergent-solubilized SecY complex from the archebacterium Methanococcus jannaschii (van der Berg et al., 2004).However, in E. coli only SecY and SecE subunits are required for cell viability while SecG is not.
In the chloroplast, homologues to SecA (cpSecA), SecY (cpSecY) and SecE (cpSecE) have been identified (Yuan et al., 1994;Roy & Barkan, 1998;Schuenemann et al., 1999), while a homologue to SecB has not been revealed.However, similar to bacteria, transport by the chloroplast Sec translocon requires the transported protein substrates to be in an unfolded conformation (Hynds et al., 1998;Marques et al., 2004).Because in the chloroplast the SecB homologue is absent, another as yet unknown stromal factor must be involved in stabilization of preprotein substrates in an unfolded state.In this context it seems that in the chloroplast the cpSec pathway in the cpSecA-cpSecYE configuration appears to operate much like the bacterial system in which the channel formed by SecYE is a passive conduit for unfolded protein substrates and the SecA ATPase is a pushing translocation motor.
The chloroplast-encoded cytochrome b 6 is an integral protein which consists of four transmembrane helices (A,B,C,D) that produce a proper environment for incorporation of two b-type hemes.Like many other integral proteins, cytochrome b 6 operates with an uncleaved signal for insertion into the biological membrane (Zhang et al., 2001).We decided to study this uncleaved signal for the insertion of cytochrome b 6 into a bacterial membrane by expressing the constructed cytochrome deletion mutants in E. coli.In a previous paper (Kroliczewski et al., 2005) we showed that the fusion of cytochrome b 6 to MBP directs the cytochrome to the Sec-dependent pathway and the cytochrome b 6 has a native structure in the bacterial cytoplasmic membrane with both NH 2 and COOH termini on the same, periplasmic side of the membrane.Dalbey and Kuhn have proposed that insertion of membrane proteins by Sec translocase requires that each hydrophobic anchor region is sensed during transport, further translocation of the protein is halted, and the anchor region is laterally released from the central channel region (Dalbey & Kuhn, 2004).
The data presented herein indicate that the insertion of mature cytochrome b 6 or its constructed deletion mutants into the cytoplasmic bacterial membrane is absolutely dependent on the presence of an N-terminal presequence artificially added which directs the protein to the E. coli membrane-insertion pathway.We also show that the proper folding and anchoring of the constructs into the bacterial membrane depends on its N-terminal part bearing helices A and B (first about 110 amino acids) and the incorporation of hemes is possible even into a truncated form of cytochrome b 6 , as long as helices B and D bearing the axial ligands to heme are present.

MATERIALS AND METHODS
Prediction of signal sequences (SSs) and signal anchor sequences (STSs) in apocytochrome b 6 .The amino-acid sequence of apocytochrome b 6 from spinach was used for the prediction of signal sequences (SSs) by the network tool SignalP ver.3.0 (http://www.cbs.dtu.dk/services/SignalP).SignalP ver.3.0 uses neural networks (NN) and hidden Markov models (HMM) to predict "secretory proteins" through analysis of protein N-terminal sequence (Nielsen & Krogh, 1998).These programs are among the most accurate methods for "secretory protein" prediction (Nielsen et al., 1999;Emanuelsson & von Heijne, 2001;Klee & Ellis, 2005) and the programs' HMMs have an uncommon ability to discriminate N-terminal SSs from N-terminal signal anchor sequences (STSs).
Construction of derivatives of the expression plasmid pMalp2b6.The recombinant plasmid pMal-c2b6 containing the gene (petB) for spinach apocytochrome b 6 (Kroliczewki & Szczepaniak, 2002) was used as a template for the creation of apocytochrome b 6 derivatives (Fig. 1).A fragment containing the appropriate part of the petB gene was amplified in each case from the plasmid using two primers.One of the primers was designed to introduce an XmnI restriction site 5′, and the second one to introduce a BamHI restriction site 3′ to the petB fragment.The primers were designed from the known sequence of petB gene from the chloroplast DNA of spinach (Heinemeyer et al., 1984).The PCR products were cleaved by double digestion with XmnI (Promega) and BamHI (Fermentas) and purified using a PCR purification kit (Macherey-Nagel).The purified products were cloned into the XmnI and BamHI sites of the pMal-p2 expression vector (New England Biolabs) by a standard T4 DNA ligase procedure.Resulting plasmids (Fig. 1) pMal-p2ABC, pMal-p2AB, pMal-p2A, pMalp2BCD, pMal-p2CD, pMal-p2C, pMal-p2BC and pMal-p2B were checked by restriction analysis and sequenced.Expression of the constructed vectors led to production of intact cytochrome b 6 or its fragments as fusion proteins with MBP.Between MBP and the cytochrome b 6 fragments an amino-acid sequence recognized by factor Xa was localized.
Construction of expression plasmids pET25b6 and pET25CD.Spinach apocytochrome b 6 petB gene or the part of the gene encoding helices C and D were amplified by PCR from isolated spinach chloroplast DNA (Herrmann, 1982).The forward primers 5′-GGCAGTGCCCATGGGTGTATCTTAC-CGGTGGGT-3′ and 5′-CATGCCATGGATGTT-GACTTGGGTTACAGGC-3′ carrying the NcoI site (underlined) for the amplification of apocytochrome b 6 or the part encoding helices C and D (respectively) and the reverse one carrying BamHI site (underlined) 5′-GGCAGTGCGGATCCCTATAAGGGACCAGAAA-TA-3′ were designed from the known sequence of the petB gene from the chloroplast DNA of spinach (Heinemeyer et al., 1984).The amplified PCR products were purified, digested and then cloned into the pET25b vector (Novagen) treated with NcoI and BamHI restriction enzymes.The resulting plasmids (Fig. 1) pET25b6 and pET25CD were checked by restriction analysis.Expression of the constructed vectors led to the production of intact cytochrome b 6 or its fragment as fusion proteins with SSpelB.
Expression of recombinant apocytochrome b 6 in E. coli CM 124-SecE depletion mutant.The SecE-depletion strain CM 124 cells (Traxler & Murphy, 1996) were transformed with pMal-p2b6 (for cytochrome b 6 expression as a fusion protein with MBP) or pET25b6 (for cytochrome b 6 expression as a fusion protein with SSpelB) and grown aerobically at 37 °C in M63 minimal medium supplemented with 0.4 % d-glucose or 0.2 % larabinose and 100 mg/ml ampicillin.For all experiments, 2 ml of overnight culture was pelleted, washed once with M63 medium and diluted 1 : 20 into 20 ml per 125 ml flask to obtain mid-logarithmic phase growth.To deplete the cells of SecE, cells were grown in the absence of larabinose.The production of the fusion protein (MBPapocytochrome b 6 or SSpelB-apocytochrome b 6 ) was induced by the addition of IPTG to a final concentration of 0.3 mM.After shaking for 1.5 h, cells were harvested by centrifugation (4 000 × g, 10 min, 4°C), resuspended in 15 ml of 30 mM Tris/HCl, pH 8.0 buffer containing 20 % sucrose and 1 mM EDTA, then incubated for 10 min at room temp.with gentle shaking.The culture was centrifuged at 8 000 × g for 10 min at 4 °C.The resulting pellet was resuspended in 10 ml of ice-cold 5 mM MgSO 4 , shaken for 10 min at 4 °C, and centrifuged as above.The obtained supernatant was the cold osmotic shock fluid.The resulting pellet of E. coli was frozen at -20 °C and used for further membrane and cytoplasmic protein preparation.For overexpression of functional SecE, l-arabinose was added 1.5 h after IPTG induction and the cultures were grown for the next 30 min, or l-arabinose was added together with IPTG (control expression).
Membrane preparation from E. coli cells.After thawing the frozen E. coli cells, the suspension (30 ml) was sonicated on ice (5 × 2 min; 70 W with 20 kHz homogeneous sound 3 mm diameter microtip sonifier cell disrupter, Sonopuls HD 2070, Bandelin) and the lysate was centrifuged at 27 000 × g for 20 min at 4 ºC.The supernatant was centrifuged again at 27 000 × g for 20 min at 4 ºC to pellet inclusion bodies debris, and the resulting supernatant was centrifuged for a further 2 h at 70 000 × g at 4 ºC.A suspension of the membrane sediment in TEN buffer was centrifuged again at 70 000 × g for 90 min at 4 ºC.The inclusion bodies and membrane fractions were examined by SDS/PAGE (Betton & Hofnung, 1996) and Western blotting.
Preparation of spheroplasts.E. coli carrying pMal-p2b6 or pMal-p2b6 derivative plasmids were grown in a standard LB medium containing 100 µg/ml of ampicillin.Induction was initiated at OD 560 about 0.6 by the addition of IPTG to the final concentration of 0.5 mM.After 4 h of shaking at 37 ºC, cells were carefully harvested by centrifugation (3 000 × g, 10 min, 4 ºC) and spheroplasts were prepared.Cell pellets were resuspended in buffer I (10 mM Tris/HCl, pH 7.5, 10 mM EDTA, 5 mM MgSO 4 , 0.7 M sucrose, 1 mM PMSF) and centrifuged at 5 000 × g for 10 min at 4 ºC.The pellets were resuspended in 10 ml of ice-cold 5 mM MgSO 4 and incubated for 10 min on ice, centrifuged as above, then resuspended once more in buffer I containing 0.2 mg/ml of lysozyme and incubated at room temp.for 25 min.The sample was centrifuged at 5 000 × g for 10 min at 4 ºC and the supernatant was removed.The pellet was resuspended in buffer I and washed twice.The fresh spheroplasts were used for agglutination tests.
Agglutination of spheroplasts.The spheroplasts were incubated with an antibody against MBP or against the COOH-terminal decapeptide (Ile 206 to Leu 215) of cytochrome b 6 (Szczepaniak & Cramer, 1990).The agglutination was analyzed by optical microscopy, immediately after the addition of the antibody.
Spectra measurements.Visible absorption spectra were recorded at room temp. in 1 cm optical-path length cuvettes (1 ml) using an M40 spectrophotometer (Carl Zeiss Jena) with a spectral bandwidth of 1 nm and a scan speed of 300 nm/min.Spectra were usually recorded in 50 mM Tris/HCl, pH 8.0 buffer containing 50 mM NaCl, 0.1 % SDS and 10 % glycerol.The samples were oxidized by adding 100 mM ferricyanide and reduced by the addition of 100 mM dithionite.
Measurements of protein concentration.MBP concentration was determined directly from the absorbance at 280 nm with an absorbance coefficient of e = 68 750 M -1 cm -1 (Betton & Hofnung, 1996).The concentration of fused protein was determined by using the BCA reagent (Sigma) according to the manufacturer's instructions with BSA as a standard.
Proteolysis of spheroplasts.Spheroplasts (50 µl, 5 mg protein/ml) were incubated at room temp.T with factor Xa for 45 min at a w/w ratio of 0.5 % of the amount of fusion protein.Cleavage was carried out in 10 mM Tris/HCl, pH 8.0 buffer containing 10 mM NaCl, and 0.05 % SDS.The reaction with factor Xa was terminated by the addition of PMSF to a final concentration of 5 mM.The spheroplasts were pelleted, washed twice with buffer I (see section: Preparation of spheroplasts), resuspended in the same buffer and used for agglutination assay.
Cleavage with factor Xa. Cleavage with factor Xa of bacterial membranes containing fusion protein was carried out at a w/w ratio of 0.4 % of total protein.The reaction mixtures in 20 mM Tris/HCl, pH 8.0 buffer containing 100 mM NaCl, 2 mM CaCl 2 and 0.03 % SDS were incubated for 10.5 hours at room temp.The reaction was blocked by the addition of PMSF to a final concentration of 5 mM.The samples containing bacterial membranes were ultracentrifuged at 70 000 × g for 2 h at 4 ºC.The pellets were resuspended in 1 ml of 20 mM Tris/HCl, pH 8.0, then precipitated with TCA (using 36 % TCA solution to final concentration of 6 %) and subjected to SDS/PAGE followed by Western blotting.In the case of dissolved fusion proteins after cleavage with factor Xa the separation of protein from MBP was carried out as described by Kroliczewski & Szczepaniak (2002).
SDS/PAGE and Western blotting.Proteins were separated on 15 % polyacrylamide gel containing 8 M urea.Denaturation of samples was achieved by mixing the protein sample or bacterial spheroplast pellets with solubilization buffer (300 mM Tris/HCl, pH 6.8, 60 % glycerol, 12 mM EDTA, 12 % SDS, 864 mM 2-mercaptoethanol, 0.05 % bromophenol blue) in a v\v ratio of 5 : 1.For membrane proteins, 2-mercaptoethanol was replaced with 480 mM DTT in a solubilization buffer.Western blotting was carried out as described by Szczepaniak & Cramer (1990).

The role of SecE in apocytochrome b 6 insertion into bacterial membrane
In a previous paper (Kroliczewski et al., 2005) we showed that the fusion of apocytochrome b 6 to MBP directed the apocytochrome to the Sec-dependent pathway.
In additional experiments, we examined the dependence of apocytochrome b 6 and fusion apocytochrome b 6 insertion on the SecE component of the secretion apparatus (Sec) of E. coli.In this case, we used the E. coli CM124 strain in which SecE can be depleted efficiently.In this strain, expression of the secE gene is under the control of the arabinose-inducible araBAD promoter.In the presence of glucose and absence of arabinose, SecE is not expressed.
In our experiment CM124 cells carrying pMal-p2b6 or pET25b6 (Fig. 1) were grown in the absence or presence of arabinose, then the insertion of MBP-apocytochrome b 6 or SSpelB-apocytochrome b 6 fusions into the cytoplasmic bacterial membrane was investigated (Fig. 2).When SecE was depleted (arabinose was not present in the medium) both MBP-apocytochrome b 6 and SSpelBapocytochrome b 6 fusions were found exclusively in the cytoplasmic fraction (Fig. 2AB, lane 3).On the other hand, in the presence of arabinose, expression of the SecE gene led to the insertion of the investigated fusions into the bacterial membrane (Fig. 2AB, lane 6).
The results presented here show that the insertion of the polytopic apocytochrome b 6 fused to MBP or SSpelB into the cytoplasmic membrane is SecE-dependent.

Prediction of signal peptides and signal anchor sequences in apocytochrome b 6
In a search for potential signal sequences and signal anchor sequences in the apocytochrome b 6 sequence, we decided to carry out a computational analysis.Most prediction programs predict "secreted proteins" by identifying an N-terminal signal sequence, a signal sequence cleavage site or a combination of both features in a target sequence.In our case, for the prediction of signal sequence (SS) and/or anchor signal sequence (STS) we chose the program SignalP ver. 3. SignalP uses two HMMs, one that models a signal peptide and the second that models a signal anchor.N-terminal signal peptide is a short polypeptide (20-25 amino-acid residues) and has no strongly conserved sequence motifs but includes three distinct sequential regions: the n (N-terminal)-region, the h (hydrophobic)-region and the c (C-terminal)region (von Heijne, 1985;1986).The signal peptide model contains submodels that describe each of these three regions.The signal anchor model contains two submodels that characterize its n-region and h-region.In the signal peptide model the h-region is limited to between six and twenty residues, the n-region must have at least one residue (and start with a methionine), and the c-region must have at least three residues.The n-region and c-region contain self-cyclic states with exponentially decaying transitions.This type of transition state allows the model to fit signal peptides possessing n-regions and c-regions with variable lengths, while still constraining the system, preventing unusually long region lengths, and thereby encapsulating the known properties of these regions.In the signal anchor module, the architecture of the n-region is the same as in the signal sequence module, but the h-region also possesses a self-cyclic, exponentially decaying transition state (Nielsen et al., 1997;Nielsen & Krogh, 1998;Klee & Ellis, 2005).
Signal anchors are also referred to as uncleaved signal peptides.However, they often have sites similar to signal peptide cleavage sites behind their hydrophobic (transmembrane) region.Therefore, a prediction method can easily be expected to mistake signal anchors for signal peptides.It is worth noting that most of eukaryotic signal anchor sequences (60 %) have mean S-scores from 0.35 to 0.70 and, in the case of protein signal sequences, the mean S-scores are greater than 0.65, while non-signal sequences have mean S-scores from 0.0 to 0.25.The aforementioned data indicate that the mean Sscore for signal anchors shows some overlap with the signal peptide distribution; however, signal anchors are generally significantly longer than signal peptides.Table 1 shows the results of the signal anchor prediction made by SignalP ver. 3 for apocytochrome b 6 deletion mutants (a maximum of the first 70 amino acids of each constructed mutant sequence were used for prediction).
For non-secretory proteins, all the S-scores represented in the SignalP ver. 3 output should ideally be very low.The HMM reports the probability of a signal anchor.The cleavage site is assigned by the probability score together with the scores for every amino acid in the n-region, h-region and c-region of the signal peptide, if one is found.Not all secretory proteins carry signal peptides or similar targeting signals.Some proteins enter the nonclassical secretory pathway without any currently known sequence motif (Bendtsen et al., 2005).To determine the subcellular localization of the expressed proteins, periplasm, cytoplasm and membranes were isolated from the induced cells.The protein content of these fractions was investigated by Western blotting.The subcellular localization of the expressed deletion mutants, MBP-BCD and MBP-BC, revealed that both fusion proteins were exclusively found in the bacterial membrane (Fig. 4A, lane 4 and Fig. 4B, lane 1).A subcellular analysis of the mutants MBP-ABC, MBP-AB and MBP-A was carried out as well (not shown) and they were detected only in the membrane fractions (Fig. 5).

Expression of apocytochrome b 6 mutants in E. coli
In contrast, the MBP-CD and MBP-C mutant fusion proteins were found exclusively in the periplasm (Fig. 6AB).To make sure that the transport of truncated apocytochrome b 6 containing only C and D transmem-  brane helices to the periplasm is not caused by MBP, we constructed a new expression vector, pET25CD.In this plasmid, the truncated apocytochrome b 6 containing only C and D transmembrane helices was fused to pelB signal sequence (SSpelB).The fusion to SSpelB directs truncated apocytochrome b 6 to the Sec-dependent pathway -similar to MBP in expression vectors derived from pMal-p2.The results of the expression of pET25CD vector revealed that the mutant protein was also found exclusively in the periplasm (not shown).

Determination of apocytochrome b 6 mutant topology in the bacterial inner membrane
To determine the topology of all fusions, E. coli cells bearing the appropriate plasmid were induced and then converted to spheroplasts.Next, the obtained spheroplasts were incubated either with an antibody against MBP or with an antibody against COOH-terminal decapeptide of cytochrome b 6 .As a control, we used uninduced or induced cells carrying pMal-p2b6 or pMal-c2b6, respectively (negative controls), and induced cells carrying Malp2b6 plasmid (positive control).As an additional (negative/positive) control, we used induced cells carrying pMal-p2b6 converted to spheroplasts followed by factor Xa treatment.The results are presented in Table 2.As it is shown, after the induction of pMal-p2A, pMal-p2B, pMal-p2AB, pMal-p2BC or pMal-p2BCD vectors and incubation of spheroplasts with an antibody against MBP, a strong or medium agglutination of the spheroplasts was observed.Such results clearly demonstrate the periplasmic location of MBP and anchoring of the studied apocytochrome b 6 deletion mutants into the bacterial cytoplasmic membrane.
Only spheroplasts derived from cells bearing the pMal-p2b6 plasmid have the possibility of showing agglutination after incubation with antibody against COOH-terminal decapeptide of cytochrome b 6 .The ob-served lack of this type of agglutination implies that the COOH terminus of apocytochrome b 6 mutants is located on the cytoplasmic side of the membrane.Additional experiments on spheroplasts bearing MBP-BCD fusion and treated with factor Xa show a lack of agglutination with an antibody against MBP.Such results clearly demonstrate that the NH 2 terminus of the studied apocytochrome b 6 deletion mutant is on the periplasmic side of the bacterial membrane, as predicted (Fig. 3).

Heme incorporation
It has been shown previously that in the E. coli cytoplasmic membrane apocytochrome b 6 and exogenic heme added to the culture medium spontaneously form a complex with similar spectroscopic properties to native cytochrome b 6 (Kroliczewski et al., 2005).Among the apocytochrome b 6 mutants only MBP-BCD has the possibility to bind hemes because of possessing the B and D helices bearing the axial ligands to heme.Indeed, when heme was present in the culture media, during the protein expression from the pMal-p2BCD vector, the dithionite-reduced spectra of isolated cytoplasmic membrane demonstrated the presence of b-type heme.After cutting off and isolating MBP, the absorption spectra of the reduced form revealed the characteristic Soret (γ-band) at 428 nm and prominent β and α bands at 532.5 nm and 562.5 nm, respectively (Fig. 7), and were similar to those of isolated cytochrome b 6 (Stuart & Wasserman, 1973;Boronowsky et al., 2001).These results demonstrate the presence of properly reconstituted hemes in the protein structure of the BCD deletion mutant and also suggest proper topology of the BCD mutant, with the NH 2 -terminus on the periplasmic and the COOH-terminus on the cytoplasmic side of the membrane, as predicted (Fig. 3).

DISCUSSION
It appears that in E. coli transport across or into the bacterial membrane is carried out by at least four independent precursor-specific transport pathways: the SRPdependent, TAT/DpH-dependent, spontaneous and Secdependent pathways (Froderberg et al., 2003;du Plessis et al., 2006).The fusion of polytopic cytochrome b 6 to   The endosymbiont hypothesis has spurred parallels between the translocation or insertion of proteins through or into E. coli cytoplasmic membrane and the translocation or insertion of chloroplast proteins through or into thylakoid membranes (Schatz & Dobberstein, 1996;Dalbey & Kuhn, 2004).This encouraged us to use E. coli for studies of cytochrome b 6 incorporation into the biological membrane.In a previous paper (Kroliczewski et al., 2005), we showed that the topogenic signals in the amino-acid sequence of the nascent chain of the chloroplast cytochrome b 6 protein were recognized by the E. coli SecYEG translocon, leading to the integration of this protein into the bacterial inner membrane.Cytochrome b 6 has a native structure in the bacterial cytoplasmic membrane but has an opposite orientation compared to that in the thylakoid (Szczepaniak & Cramer, 1990;Kurisu et al., 2003).Recently it has been reported that polytopic cytochrome b 6 (which naturally contains four transmembrane helices: A,B,C,D) can spontaneously insert into bacterial cytoplasmic membrane (Dreher et al., 2007).Such a conclusion is in opposition to our earlier findings.That is why we decided to perform additional experiments to verify the inconsistent data.We took up the study with E. coli CM124 cells in which SecE (integral subunit of SecYE translocon) could be depleted.The obtained results show that apocytochrome b 6 expressed in cells with depleted SecE is found only in the cytoplasm (Fig. 2AB, lane 3), with no signal deriving from apocytochrome b 6 detected in the membrane fraction (Fig. 2AB, lane 5).However, expression in the presence of arabinose leads to incorporation of apocytochrome into the bacterial membrane (Fig. 2AB, lane 6).These results clearly demonstrate that the insertion of apocytochrome b 6 into the membrane is strongly dependent on SecE and no spontaneous incorporation was observed.So, the insertion of polytopic cytochrome b 6 into the cytoplasmic bacterial membrane is absolutely dependent on the presence of an N-terminal presequence artificially added, and its N-terminus is always transported to the periplasmic side of the bacterial membrane (Kroliczewski et al., 2005).Here arises the question: where does from the mentioned divergence of the results obtained by Dreher et al. (2007) and our group come?In the cited paper the presence of cytochrome b 6 in the membrane was indicated only by measurements of visible spectra of membranes isolated from E. coli.From the literature it is known that exogenic heme can interact with the membrane itself or via other proteins in a non-specific manner (Cassoly, 1978;Marden et al., 1989;Solar et al., 1989;Solar & Shaklai, 1989;Leclerc et al., 1993).Such unspecific interaction of heme with bacterial membrane was also observed by us in a previous study (Kroliczewski et al., 2005).In this context, the results presented by Dreher et al. (2007) should be verified for example in washing experiments of isolated bacterial membrane with chaotropic agents (in such conditions integral proteins are not washed out from the membrane) or in experiments with spheroplasts (prepared from induced bacterial cells) using proper antibodies/proteases which could allow the study of the transmembrane topology of expressed cytochrome.Such experiments were not done in the research in question.Moreover, the visible absorption spectra of reduced cytochrome b 6 reveal specific bands characterized as α, β and γ, named the Soret band.In the discussed paper, visible spectra of reduced membranes isolated from bacteria do not reveal a characteristic well-marked Soret band.Additionally, to confirm that visible spectra are characteristic for cytochrome b 6 it is recommended to analyze the width of the α-band at the half-height; such analysis was not carried out either.In light of the presented facts, in our opinion the authors (Dreher et al., 2007) could observe signals deriving from hemes interacting with the bacterial membrane in a nonspecific manner.
Returning to the translocation/insertion of proteins across/into the membrane, it has been proposed that when the elongating sequence passes through the SecY complex, the translocon has to "decide" whether the segment should be secreted or shunted into the membrane as a TM helix.Their integration into the membrane utilizes a mechanism that requires an upstream cleavable SS or noncleavable TM segment.The subsequent appearance of the signal anchor sequence (STS) within the   translocation apparatus aborts the SS/TM-initiated translocation of the polypeptide chain across the membrane.STSs occur as hydrophobic segments, flanked by charge residues, which are sufficiently long to form TM α-helix and provide a permanent anchorage (Saaf et al., 1998).
Other investigations concerning the insertion of proteins into the membrane have shown that the coexpression of two or three complementary polypeptide fragments of rhodopsin allows the formation of noncovalently linked rhodopsin, showing that the covalent linkage of the polypeptide chain is not required for the in vivo insertion into the membrane, folding and assembly of rhodopsin (Ridge et al., 1995).It has also been shown that many rhodopsin fragments with discontinuities in the intradiscal, transmembrane and cytoplasmic regions fold and insert into the membrane (Ridge et al., 1996).
During our study on cytochrome b 6 assembly into the membrane, the following question arose: which part/parts of cytochrome b 6 apoprotein is/are responsible for the protein insertion into the membrane?In the case of polypeptide fragments of cytochrome b 6 fused to MBP expressed in E. coli only TM helices A and B and constructs containing both these helices were folded and inserted into the membrane (Fig. 5).Unexpectedly, constructed mutants containing TM helix C or helices CD were transported through the cytoplasmic membrane and were found exclusively in the bacterial periplasm (Fig. 6AB), implying that helices C and D do not include an STS.Computational analysis has shown a lack of STS in fragments containing helix C or D and simultaneously revealed the presence of STS in helices A and B (Table 1); these predictions are in agreement with our experimental results.In contrast to rhodopsin (Ridge et al., 1995;Saaf et al., 1998), for proper folding and correct insertion into the membrane (with the possibility of heme insertion into the protein structure) helices C and D of cytochrome b 6 require covalent linkage to the preceding TM helix B, and the presence of helix A is not necessary for the anchoring process (Fig. 4 and Fig. 7).
Other important results obtained in our study are that the spectral and redox properties of the BCD mutant of cytochrome b 6 mimic those of the native protein and are similar to those of cytochrome b 6 spectra obtained from isolated protein or cytochrome b 6 -f complex (Stuart & Wasserman, 1973;Boronowsky et al., 2001).The spectral and redox properties of the BCD mutant are typical for bis-histidine liganded btype cytochromes.This very strongly suggests that the structural features responsible for the significant redox differences between the two hemes reside in the BCD part of cytochrome b 6 itself, rather than arising from a higher order conformation induced by the other subunits or transmembrane helix A. It is also clear that membrane-incorporated BCD helices reveal a proper structure and an appropriate distance between histidine axial ligands that allow proper heme incorporation.

CONCLUDING REMARKS
In the present study we have shown that the fused signal sequence (SS) is responsible for the orientation of apocytochrome b 6 into bacterial cytoplasmic membrane and the fused part is located on the periplasmic side.Helices A and B of apocytochrome b 6 are responsible for the anchoring of cytochrome b 6 and B (with or with-out A) ensures insertion and proper folding of helices C and D, while helices C and D do not possess a functional anchor signal sequence (STS).The results presented here also show that proper incorporation of hemes into a truncated form of apocytochrome b 6 is possible, as long as helices B and D (parts bearing the axial ligands to heme) are present.

Figure 1 .
Figure 1.Diagram of petB gene constructs used to overexpress cytochrome b 6 mutants DNA sequence corresponding to the transit peptide in pET25CD or pET25b6 is indicated by solid black box.Positions of membranespanning regions are indicated by open boxes.Numbers refer to the amino-acid sequence of apocytochrome b 6 where the initiating methionine (M) is 1.Relevant restriction sites are indicated by arrows.The deleted regions are indicated by dashed lines.

Figure 2 .
Figure 2. Effect of SecE depletion on MBP-apocytochrome b 6 and SSpelB-apocytochrome b 6 location in E. coli (A) Western blot analysis of MBP-apocytochrome b 6 fusion protein expressed in CM124 strain.Lane 1, total protein fraction from E. coli cells before induction; lane 2, total protein fraction from E. coli cells after induction; lane 3, cytoplasmic fraction; lane 4, periplasmic fraction; lane 5, membrane fraction; lane 6, membrane fraction isolated after culturing with arabinose.(B) Western blot analysis of SSpelB-apocytochrome b 6 fusion protein expressed in CM124 strain.Lane description as in panel A. For immunodetection, polyclonal antibody against COOH-terminal part of apocytochrome b 6 was used.On each lane, 20 mg of protein was applied.

Figure 1
Figure 1 illustrates the derivatives of cytochrome b 6 constructed for this study.They were made from spinach petB gene encoding cytochrome b 6 .Different regions of cytochrome b 6 were fused to the carboxyl terminus of periplasmic bacterial MBP or to the carboxyl terminus of SSpelB in an attempt to identify regions of cytochrome b 6 which function as SSs or STSs.The predicted topology and orientation of the expressed apocytochrome b 6 mutants in bacterial cytoplasmic membrane are shown in Fig. 3. Having established that apocytochrome b 6 overexpressed as a fusion protein to MBP can be incorporated into bacterial cytoplasmic membrane(Kroliczewski et al., 2005), we tested whether the removal of a specific por-

Figure 3 .
Figure 3. Expected topology and orientation of apocytochrome b 6 mutants in bacterial cytoplasmic membrane Membrane helices are named according to apocytochrome b 6 terminology.

Figure 7 .
Figure 7. Redox difference spectrum of reconsituted BCD mutant of cytochrome b 6 Ferricyanide and dithionite were used as oxidant and reductant, respectively.For details, see the text.