Lactococcus lactis cannot synthesize haem, but when supplied with haem, expresses a cytochrome bd oxidase. Apart from the cydAB structural genes for this oxidase, L. lactis features two additional genes, hemH and hemW (hemN), with conjectured functions in haem metabolism. While it appears clear that hemH encodes a ferrochelatase, no function is known for hemW. HemW-like proteins occur in bacteria, plants and animals, and are usually annotated as CPDHs (coproporphyrinogen III dehydrogenases). However, such a function has never been demonstrated for a HemW-like protein. We here studied HemW of L. lactis and showed that it is devoid of CPDH activity in vivo and in vitro. Recombinantly produced, purified HemW contained an Fe–S (iron–sulfur) cluster and was dimeric; upon loss of the iron, the protein became monomeric. Both forms of the protein covalently bound haem b in vitro, with a stoichiometry of one haem per monomer and a KD of 8 μM. In vivo, HemW occurred as a haem-free cytosolic form, as well as a haem-containing membrane-associated form. Addition of L. lactis membranes to haem-containing HemW triggered the release of haem from HemW in vitro. On the basis of these findings, we propose a role of HemW in haem trafficking. HemW-like proteins form a distinct phylogenetic clade that has not previously been recognized.

INTRODUCTION

Lactococcus lactis is a Gram-positive bacterium that is devoid of a haem biosynthetic pathway and usually does not respire. However, if supplied with protoporphyrin IX or haem b in growth medium L. lactis is capable of respiratory growth due to the presence of a terminal cytochrome bd ubiquinol oxidase [13]. L. lactis possesses a functional ferrochelatase, HemH, which allows the organism to use protoporphyrin IX, in addition to haem b, for cytochrome bd synthesis [4].

Additionally, a gene annotated as hemN is present in the genome of L. lactis. Since this annotation is most confusing in the realm of the present work, we henceforth call hemN of L. lactishemW’. HemW shares approximately 50% amino acid sequence similarity with HemN of Escherichia coli. HemN of Salmonella Typhimurium and E. coli have been shown to function as oxygen-independent CPDHs (coproporphyrinogen III dehydrogenases; previously called coproporphyrinogen III oxidases or CPO), which catalyse the oxidative decarboxylation of coproporphyrinogen III to protoporphyrinogen IX in bacterial haem biosynthesis [57]. Several observations do, however, argue against a function of L. lactis HemW as a CPDH. First, L. lactis is unable to synthesize haem; it can only respire when grown in the presence of protoporphyrinogen IX or hemin. Hence the presence of CPDH in this organism would be futile. Secondly, HemW exhibits key sequence divergence from proteins with demonstrated CPDH activity: it is missing the 47 N-terminal amino acids and the fourth cysteine residue of the conserved CX3CX2CXC motif, and it exhibits poor sequence similarity to HemN in the C-terminal 150 amino acid residues. Finally, a HemW-like protein of Synechocystis sp. PCC 6803, Sll1917, did not exhibit CPDH activity in vivo or in vitro [8].

hemW-like genes are widespread in Gram-positive and Gram-negative bacteria. In Bacillus subtilis, the two hemW-like genes have been annotated as hemN and hemZ, and have been proposed to encode CPDH. The proposal was based on the coproporphyrin III accumulation by a B. subtilis hemN hemZ double mutant, but no direct enzyme assays with purified proteins had been conducted [9,10]. Interestingly, hemW-like genes also occur in plant and animal cells, although HemN-type CPDH enzymes are believed to be a strictly prokaryotic feature. Given the unclear role of HemW-like proteins, in the present study we have investigated the function of L. lactis HemW. We found that the protein was devoid of CPDH activity, both, in vivo and in vitro. HemW could bind haem covalently and could transfer haem to a membrane component in vitro, suggesting a function of HemW as a haem chaperone.

EXPERIMENTAL

Bacterial strains and growth conditions

The E. coli strain JW3838 (ΔhemN) was obtained from the National BioResource Project. The E. coli strain W3110 was used as wild-type strain. E. coli strains were grown aerobically in LB (Luria–Bertani) medium or anaerobically (nitrifying) in LB medium containing 10 mM NaNO3. Cultures of 125 ml were inoculated with 1.5 ml of aerobic overnight cultures and grown anaerobically at 37°C. In complementation experiments, protein expression was induced with 0.1 mM IPTG (isopropyl-β-D-thiogalactoside) at the time of inoculation. Cultures were grown in bottles rotating at 120 rev./min at 37°C and growth was followed by measuring the attenuance at 600 nm. L. lactis IL1403 was grown aerobically in M17 medium [11]. Growth was monitored by measuring the attenuance at 600 nm with a Lambda 16 spectrophotometer (PerkinElmer). For complementation, JW3838 was transformed with the plasmid pET3ahemN expressing E. coli hemN [7], the control vector pProExHTa or pWN309 expressing L. lactis hemW (the present study).

Cloning of L. lactis HemW

The hemW gene was PCR-amplified from L. lactis genomic DNA using primers ha44 (5′-GTCTGGCGCCTTGCAAAAACCTAATTCGGCTTATTTTC) and ha45 (5′-ACAGTCTAGATAACTTTCACAAAAATTCATCACGATAGC), harbouring a NarI and an XbaI-restriction site (underlined) respectively. The PCR product (1187 bp) was cloned into pCRBluntII-TOPO (Invitrogen) according to the manufacturer's instructions, yielding the plasmid pHA11003. The resulting construct was digested with NarI and XbaI and the excised fragment ligated into pProExHTa (Invitrogen), previously digested with the same enzymes, yielding the plasmid pWN309. This vector encoded HemW with an N-terminal His6 tag and a cleavage site for rTEV (recombinant tobacco edge virus) protease. Cleaved HemW had the N-terminal methionine replaced by Gly-Ala.

Purification of HemW and activity measurements

His6–HemW was produced in E. coli BL21(DE3) cells containing the plasmid pWN309. Cultures of 4 litres were grown aerobically at 37°C in shaking flasks. hemW gene expression was induced at an attenuance at 600 nm of 0.7 by the addition of 0.1 mM IPTG and cultivation was continued overnight at 18°C. Cells were harvested by centrifugation at 8000 g for 15 min at 4°C, washed once with 400 ml of PBS and once with 200 ml of PBS. The resulting cell pellet was stored at −20°C. For HemW purification, cells were resuspended in 45 ml of buffer A (20 mM Tris/HCl, pH 8, and 500 mM NaCl) containing a tablet of EDTA-free complete protease inhibitor cocktail (Roche) and 10 μg/ml of DNaseI (Roche). Cells were broken by three passages through a French press at 40 MPa or with an Emulsiflex C5 homogenizer (Avestin). Cell debris and membranes were removed by ultracentrifugation at 35000 rev./min (Sorvall T-865 rotor) for 1 h at 4°C. The supernatant obtained after the ultracentrifugation was sterile-filtered through a 0.22 μm membrane filter (Millipore). The soluble protein fraction was loaded on to a His-Trap column (GE Healthcare) and His6–HemW was eluted with a gradient of 20 mM to 1 M imidazole in buffer A. HemW-containing fractions were pooled and concentrated by ultrafiltration with an Amicon membrane of 30 kDa molecular mass cut-off (Millipore). The His6 tag was cleaved off overnight at 4°C with rTEV protease (Invitrogen) according to the manufacturer's instructions, but without EDTA in the cleavage buffer. Cleaved HemW was passed through a His-Trap column and the flow-through concentrated as described above. To separate apo- from holo-HemW, preparations were resolved on Superdex S-75 26/60 columns (GE Healthcare) in 150 mM NaCl and 10 mM Tris/HCl, pH 8, at 2.5 ml/min. HemW with the His6 tag was used for the experiments shown in Figures 2, 3(A–C) and 5 and Supplementary Figure S1(A). Protein concentrations were determined with the Bio-Rad protein assay, using BSA as a standard. The iron content of HemW was assessed according to Beinert [12]. For some experiments, HemW was purified under strictly anaerobic conditions. However, this proved not to be generally necessary since the Fe–S (iron–sulfur) cluster of HemW was stable in air for several days. CPDH activity was measured as described previously [7].

Isolation of membrane vesicles

A volume of 500 ml of M17 medium in a 1 litre Erlenmeyer flask with four chicanes was inoculated at a dilution of 1:50 with an overnight culture of L. lactis and grown at 30°C with shaking at 160 rev./min until the early stationary phase. Cells were harvested at 5000 g for 10 min and washed once in 0.25 vol. of 2 mM MgSO4 and once in 0.125 vol. 2 mM MgSO4. Cells were suspended in 4 ml of 400 mM KCl, 50 mM potassium glycylglycine, 2 mM MgSO4, pH 7.2, containing 1 mM PMSF. Subsequently, 4 mg of lysozyme/mg of wet cells were added, followed by incubation with rotary shaking at 240 rev./min and 37°C for 1 h. All subsequent steps were performed on ice. DNase I (50 μg/ml) was added and the suspension was passed three times through a French press at 70 MPa. Cell debris was removed by centrifugation for 12 min at 23000 g. The supernatant was subjected to ultracentrifugation for 1 h at 42000 rev./min (Sorvall T-865 rotor) and membrane pellets were washed with 0.7 ml of 10 mM sodium-Hepes and 5% glycerol, pH 7.5, per gram of wet cell starting material. Membranes were pelleted and resuspended as before yielding washed membrane vesicles.

Absorption spectroscopy

UV–visible absorption spectra of haem and HemW–haem complexes were recorded on a Lambda 16 spectrophotometer (PerkinElmer) in 50 mM Tris/HCl, pH 7.5, and 150 mM NaCl, using the same buffer as a blank. For reduced haem spectra, 5 mM sodium dithionite was added to the sample immediately before the measurement. The following molar absorption coefficients (ϵ) were used: haem b at 415 nm, ϵ=102 mM−1·cm−1 [13]; and apo-HemW at 280 nm, ϵ=35.3 mM−1·cm−1 (calculated).

Haem staining

For haem staining, proteins were separated on standard SDS gels (12.5%) [14] followed by electrophoretic transfer on to nitrocellulose membranes. Haem was visualized by its intrinsic peroxidase activity as described previously [15], using laboratory-made chemiluminescent reagents (100 mM Tris/HCl, pH 8.5, 1.25 mM 3-aminophtalhydrazide, 0.2 mM p-coumaric acid and 0.01% H2O2).

Haem binding and affinity measurements

For preparative haem binding, holo-HemW (23 μM) or apo-HemW (6 μM) were incubated with an equimolar concentration of haem in 150 mM NaCl and 10 mM Tris/HCl, pH 8, at room temperature (22°C) for 30 min. Samples of 100 μl were then chromatographed on Superdex 75 PC 3.2/30 columns (GE Healthcare) at a flow rate of 0.05 ml/min. The absorption of the eluted proteins was measured at 280 and 410 nm. Columns were calibrated with alcohol dehydrogenase (Mr=150000), BSA (66000), carbonic anhydrase (29000), and cytochrome c (12000). To measure haem-binding affinity, 2.5–50 μM of haem was added to 4 μM HemW in 150 mM NaCl, 10 mM Tris/HCl, pH 8.0. Quenching of intrinsic tryptophan fluorescence was recorded with a Fluoromax-3 fluorometer (Horiba Scientific) by exciting the sample at 295 nm (λex) and recording the λem from 310 to 450 nm. Results were analysed as described by Tedesco et al. [16]. The fluorescence intensity F is related to the haem concentration by (F0F)/(FF)=([haem]/Kd)n, where F0 is the fluorescence intensity of the enzyme alone and F is the fluorescence intensity of HemW saturated with haem. The association constant, Ka, which is reciprocal to Kd, was obtained by plotting log [(F0F)/(FF)] against log [haem]. The area under the fluorescence spectra in the range 310–450 nm was taken as the fluorescence intensity F. The number of equivalent haem binding sites was derived from the slope of a double-logarithmic plot, whereas the value of log Kd was taken as log [haem] at [(F0F)/(FF)]=0.

In vitro haem discharge from HemW

HemW (10 μM) was loaded with haem by overnight incubation with an equimolar concentration of haem at room temperature in 50 mM Tris/HCl and 150 mM NaCl, pH 8. For haem discharge, 250 pmol of the HemW–haem complex were incubated with 50 μg of membrane vesicles and 0.5 mM NADH for 1 h at 30°C in 50 μl of the same buffer under agitation. Aliquots containing 50 pmol of HemW–haem were then resolved on 12.5% Tris-glycine SDS gels and electrophoretically transferred on to nitrocellulose membranes. Membranes were first stained for haem, followed by protein staining with Ponceau red as described above.

RESULTS

Primary structure of HemW

L. lactis HemW superficially resembles CPDH (HemN) of E. coli, the best-characterized CPDH enzyme. However, there are some important differences between these proteins. Figure 1 shows a sequence alignment of E. coli HemN with L. lactis HemW and a number of other HemW-like proteins which had (erroneously) been assigned as HemN. The N-terminal 47 amino acids of E. coli HemN are missing in HemW and HemW-like proteins. This extended N-terminus contains a domain that has been named ‘trip-wire’, P21RYTSYPTA, which appears to serve in substrate recognition and active-site closure, and thus to be important for CPDH activity [17]. Also, the fourth cysteine residue of the Fe–S cluster motif of E. coli HemN, CX3CX2CXC, is replaced by phenylalanine in HemW and related proteins. Finally, the C-terminal 150 amino acid residues of HemW-like proteins exhibit only poor sequence similarity to E. coli HemN. This divergence casts doubt on the function of HemW and related proteins as a CPDH.

Amino acid sequence comparison of E. coli HemN with HemW-like proteins

Figure 1
Amino acid sequence comparison of E. coli HemN with HemW-like proteins

The lower six sequences represent HemW-like proteins from the indicated bacteria; the sequence at the top is that of E. coli HemN, which may be the only true CPDH of this group of sequences. The four cysteine residues of the radical SAM enzyme motif of CPDH enzymes are indicated by arrows. Eco_YggW, E. coli YggW (GenBank® accession number NP_417430.1); Pae_HemN, Pseudomonas aeruginosa HemN (GenBank® accession number YP_001345880.1); Bja_HemN, Bradyrhizobium japonicum HemN (GenBank® accession number NP_767312.1); Bsu_HemN, B. subtilis HemN (GenBank® accession number NP_390428.2); Sau_HemN, Staphylococcus aureus HemN (GenBank® accession number NP_374696.1); Lla_HemN, L. lactis HemN (GenBank® accession number NP_267295.1); Eco_HemN, E. coli HemN (GenBank® accession number NP418303.2). The amino acid sequence alignment was generated with ClustalW2 and GeneDoc was used to generate the display. Amino acid numbering is given above the aligned sequences; the asterisks indicate decades.

Figure 1
Amino acid sequence comparison of E. coli HemN with HemW-like proteins

The lower six sequences represent HemW-like proteins from the indicated bacteria; the sequence at the top is that of E. coli HemN, which may be the only true CPDH of this group of sequences. The four cysteine residues of the radical SAM enzyme motif of CPDH enzymes are indicated by arrows. Eco_YggW, E. coli YggW (GenBank® accession number NP_417430.1); Pae_HemN, Pseudomonas aeruginosa HemN (GenBank® accession number YP_001345880.1); Bja_HemN, Bradyrhizobium japonicum HemN (GenBank® accession number NP_767312.1); Bsu_HemN, B. subtilis HemN (GenBank® accession number NP_390428.2); Sau_HemN, Staphylococcus aureus HemN (GenBank® accession number NP_374696.1); Lla_HemN, L. lactis HemN (GenBank® accession number NP_267295.1); Eco_HemN, E. coli HemN (GenBank® accession number NP418303.2). The amino acid sequence alignment was generated with ClustalW2 and GeneDoc was used to generate the display. Amino acid numbering is given above the aligned sequences; the asterisks indicate decades.

In the databases, there are many HemW-like proteins, but they have most of the time been annotated as HemN, less frequently as HemZ, and sometimes with a generic gene name, such as in the case of E. coli YggW. Many of these predicted proteins have been implied to function as CPDH. Only the HemW-like protein of Synechocystis sp. PCC 6803, Sll1917, has been shown to be devoid of CPDH activity in vivo and in vitro [8]. Synechocystis does harbour a CPDH activity that is encoded by the hemN-like gene Sll1185.

HemW does not exhibit CPDH activity

Since the L. lactis hemW gene was originally annotated as hemN and thus implying CPDH activity, we tested anaerobically purified HemW for CPDH activity according to Layer et al. [7]. The assay is based on the shift of the florescence maxima of coproporphyrin III at 620 nm to that of protoporphyrin IX at 634 nm; coproporphyrin and protoporphyrin are formed by oxidation of the corresponding porphyrinogens with H2O2. The assay was conducted under strictly anaerobic conditions and cell-free extracts of E. coli or L. lactis were added to supply the native, but still unknown, terminal electron acceptor for the oxidative decarboxylation of coproporphyrinogen III. The CPDH activity of E. coli HemN, used as a positive control, was readily apparent by the formation of a fluorescence peak at 634 nm due to the formation of protoporphyrinogen IX (Figure 2A). No CPDH activity could be detected for anaerobically purified HemW.

Determination of CPDH activity

Figure 2
Determination of CPDH activity

(A) Conversion of coproporphyrinogen III into protoporphyrinogen IX was determined by fluorometry as described previously [7]. Solid line, 75 μg/ml E. coli HemN; broken line, negative control without enzyme; dotted line, 75 μg/ml L. lactis HemW. (B) Complementation of an E. coli ΔhemN mutant. E. coli JW3838 (ΔhemN) was transformed with pProExHTa control vector (○), pET3ahemN expressing E. coli HemN (Δ), or pWN309 expressing L. lactis HemW (□). Wild-type E. coli was used as a control (●).

Figure 2
Determination of CPDH activity

(A) Conversion of coproporphyrinogen III into protoporphyrinogen IX was determined by fluorometry as described previously [7]. Solid line, 75 μg/ml E. coli HemN; broken line, negative control without enzyme; dotted line, 75 μg/ml L. lactis HemW. (B) Complementation of an E. coli ΔhemN mutant. E. coli JW3838 (ΔhemN) was transformed with pProExHTa control vector (○), pET3ahemN expressing E. coli HemN (Δ), or pWN309 expressing L. lactis HemW (□). Wild-type E. coli was used as a control (●).

To rule out in vitro artefacts, HemW was also tested for CPDH activity in vivo. To this end, the E. coli ΔhemN mutant JW3838 was used as a test system. Due to the lack of CPDH activity and thus haem biosynthesis, this mutant exhibits impaired growth under anaerobic conditions compared with the wild-type (Figure 2B). When this ΔhemN mutant was transformed with the E. coli hemN-expressing plasmid pET3ahemN, the phenotype was reversed to wild-type. In contrast, plasmid pWN309, which expresses L. lactis HemW at high levels in E. coli, did not complement the growth deficit of the E. coli ΔhemN mutant. This observation further supports the hypothesis that HemW does not exhibit CPDH activity.

These results are also in agreement with the reported absence of CPDH activity in a Salmonella Typhimurium mutant with an intact hemW-like gene, yggW, but lacking functional hemN and hemF genes. hemN and hemF encode oxygen-independent and the oxygen-dependent CPDH activity respectively [5]. Taken together, these findings argue against a function of HemW of L. lactis and related proteins as CPDH enzymes.

Dimeric HemW contains an Fe–S cluster

HemW features a CX3CX2C motif, which is a typical feature of many Fe–S proteins (Figure 1). Indeed, HemW which was purified under anaerobic conditions exhibited a brown colour. Absorption spectroscopy revealed an absorption maximum at 408 nm (Supplementary Figure S1A at http://www.BiochemJ.org/bj/442/bj4420335add.htm), typical of Fe–S clusters and almost identical with the absorption spectrum of HemN. By quantitative iron determination, purified HemW was found to contain 2.5±0.5 Fe per monomer, again identical with E. coli HemN, confirming the presence of an Fe–S cluster. It is typical for recombinantly produced Fe–S cluster proteins that not all proteins carry intact clusters, which accounts for the iron amount measured. Further characterization of the Fe–S cluster by Mössbauer and EPR spectroscopy will be the subject of future studies.

Size-exclusion chromatography of purified HemW yielded two main species with relative molecular masses corresponding to monomeric and dimeric HemW (Figure S1B). Only dimeric HemW contained the Fe–S cluster, whereas monomeric HemW was Fe–S cluster-free and colourless. The 4Fe–4S cluster of HemN of E. coli was characterized by X-ray crystallography and was found to be coordinated by only three of the four cysteines of the CX3CX2CXC motif [17]; since these three cysteines are conserved in HemW, it conceivably also contains a (4Fe–4S) cluster in the native state. Whatever the nature of the Fe–S cluster in HemW is, it appears to be important for the dimerization of the protein.

HemW binds haem in vivo and in vitro

In light of the fact that L. lactis does not synthesize haem, but is able to take it up from the growth medium and to incorporate it into cytochromes, we hypothesized that L. lactis HemW is involved in haem uptake or trafficking. To this end, we tested whether HemW is a haem-binding protein. When produced as a recombinant protein in E. coli, HemW was primarily localized in the cytosol and was haem-free (Figure 3A). A minor fraction of the HemW was, however, associated with the E. coli cytoplasmic membrane. Tightly bound haem could be detected in the membrane-associated HemW by haem-staining following SDS/PAGE and Western blotting.

Haem-binding by HemW

Figure 3
Haem-binding by HemW

(A) Covalent binding of haem by HemW in vivo. From E. coli producing L. lactis HemW, 15 μg each of total lysate (T), cytosolic fraction (C), and membrane fraction (M) were resolved by SDS/PAGE, blotted on to nitrocellulose membranes and stained for haem and protein. Covalent HemW–haem complex formed in vitro in the presence of dithionite was used as a control (Ctrl). The molecular mass in kDa is given between the two panels. (B) Effect of reductants on haem binding to HemW in vitro. Purified HemW (10 μM) was incubated overnight with an equimolar amount of haem. To 5 μM of the formed haem–HemW complex either no (Ctrl) or 0.5 μM of the reducing agents indicated in the Figure were added and incubated for 1 h at 30°C under agitation. Haem and protein were visualized as in (A). DTN, dithionite. (C) Covalent binding of haem by HemW in vitro. Purified proteins (15 μM) were incubated with a two-fold molar excess of haem for 65 h at 25°C in the presence of 5 mM sodium dithionite. Samples were resolved by SDS/PAGE, blotted on to nitrocellulose membranes and stained for haem. HemN, 75 pmol of E. coli HemN; HemW, 75 pmol of L. lactis HemW; Lys, 75 pmol of lysozyme; Cytc, 5 pmol of untreated horse heart cytochrome c. The molecular mass in kDa is given on the left-hand side. (D) Absorption spectra of the HemW–haem complex. Spectra were recorded with 10 μM HemW incubated with an equimolar amount of haem under reducing conditions (solid line) or in air without a reducing agent (dotted line). The absorption spectrum of free haem is also shown (dashed line).

Figure 3
Haem-binding by HemW

(A) Covalent binding of haem by HemW in vivo. From E. coli producing L. lactis HemW, 15 μg each of total lysate (T), cytosolic fraction (C), and membrane fraction (M) were resolved by SDS/PAGE, blotted on to nitrocellulose membranes and stained for haem and protein. Covalent HemW–haem complex formed in vitro in the presence of dithionite was used as a control (Ctrl). The molecular mass in kDa is given between the two panels. (B) Effect of reductants on haem binding to HemW in vitro. Purified HemW (10 μM) was incubated overnight with an equimolar amount of haem. To 5 μM of the formed haem–HemW complex either no (Ctrl) or 0.5 μM of the reducing agents indicated in the Figure were added and incubated for 1 h at 30°C under agitation. Haem and protein were visualized as in (A). DTN, dithionite. (C) Covalent binding of haem by HemW in vitro. Purified proteins (15 μM) were incubated with a two-fold molar excess of haem for 65 h at 25°C in the presence of 5 mM sodium dithionite. Samples were resolved by SDS/PAGE, blotted on to nitrocellulose membranes and stained for haem. HemN, 75 pmol of E. coli HemN; HemW, 75 pmol of L. lactis HemW; Lys, 75 pmol of lysozyme; Cytc, 5 pmol of untreated horse heart cytochrome c. The molecular mass in kDa is given on the left-hand side. (D) Absorption spectra of the HemW–haem complex. Spectra were recorded with 10 μM HemW incubated with an equimolar amount of haem under reducing conditions (solid line) or in air without a reducing agent (dotted line). The absorption spectrum of free haem is also shown (dashed line).

When haem-free, purified HemW was incubated with haem in vitro, it also bound haem tightly (Figure 3B), suggesting that the haem-free state of HemW was due to a lack of haem during overexpression. Dithionite strongly enhanced haem binding, GSH enhanced it slightly and NADH had no effect on haem binding, e.g. there was the same small, but significant, amount of haem binding to HemW observed in the absence of reductant. No stable haem binding was observed for two control proteins treated under identical conditions, namely HemN from E. coli and chicken egg white lysozyme (Figure 3C). The HemW–haem complexes exhibited a slightly decreased electrophoretic mobility compared with haem-free HemW. The great stability of the HemW–haem complex suggested that the haem was bound covalently.

The absorption spectrum of the HemW–haem complex, generated by incubation with equimolar amounts of haem in the presence of 5 mM sodium dithionite, exhibited absorption maxima at 425, 529.5 and 559 nm (Figure 3D). These are typical absorption features of b-type cytochromes in the reduced state that have been reported for E. coli ‘cytochrome b1’ (bacterioferritin) [18] and other b-type cytochromes [19]. Interestingly, the absorption spectrum of the HemW–haem complex closely resembled the spectra of CcmE–haem and CcmF–haem (Ccm is cytochrome c maturation protein) of E. coli, two proteins that function as haem chaperones in cytochrome c maturation [2022].

L. lactis HemW contains a single tryptophan residue in the sequence context HNXXYW, a motif that is conserved in HemW-like proteins. The intrinsic fluorescence of this tryptophan residue was quenched upon haem binding. This allowed the determination of the haem-binding affinity of HemW by fluorescence quenching (Figure 4A). Haem bound to a single binding site with a KD value of 8±0.6 μM under non-reducing conditions, as assessed in three independent experiments.

Haem binding and oligomeric state of HemW

Figure 4
Haem binding and oligomeric state of HemW

(A) Measurement of haem-binding affinity. HemW (4 μM) was incubated without (continuous line) or with 2.5, 5, 10, 20 and 25 μM haem (broken lines from top to bottom) and the intrinsic tryptophan fluorescence was monitored. Saturation was reached at 50 μM haem (dotted line). Inset: double-logarithmic plot of HemW fluorescence as a function of the haem concentration. (B) Analysis of the oligomeric state. Monomeric, iron-free HemW (upper panel), or dimeric Fe–S cluster-containing HemW (lower panel) was incubated with an equimolar amount of haem, followed by analytical gel filtration. Continuous line, protein absorption at 280 nm; broken line, haem absorption at 410 nm. Other details of the procedure are described in the Experimental section.

Figure 4
Haem binding and oligomeric state of HemW

(A) Measurement of haem-binding affinity. HemW (4 μM) was incubated without (continuous line) or with 2.5, 5, 10, 20 and 25 μM haem (broken lines from top to bottom) and the intrinsic tryptophan fluorescence was monitored. Saturation was reached at 50 μM haem (dotted line). Inset: double-logarithmic plot of HemW fluorescence as a function of the haem concentration. (B) Analysis of the oligomeric state. Monomeric, iron-free HemW (upper panel), or dimeric Fe–S cluster-containing HemW (lower panel) was incubated with an equimolar amount of haem, followed by analytical gel filtration. Continuous line, protein absorption at 280 nm; broken line, haem absorption at 410 nm. Other details of the procedure are described in the Experimental section.

Haem binding does not influence the oligomeric state of HemW

Purified HemW had been found to consist of a mixture of Fe–S cluster-containing dimers and iron-free monomers (see above). To test whether both forms of HemW were able to bind haem, they were separated by gel permeation chromatography and individually incubated with equimolar amounts of haem, followed by analytical gel filtration. The haem content of the fractions was measured by their absorption at 410 nm (Figure 4B). The two forms of HemW became haem-loaded to similar extents, namely 17 and 23% for iron-containing HemW and monomeric iron-free HemW respectively. Both forms of HemW also retained their respective oligomeric state upon haem binding. Thus the Fe–S cluster of HemW is not required for haem binding and haem binding does not alter the oligomeric state of HemW.

NADH-dependent transfer of haem from HemW to a membranelocalized target

In many bacteria, hemin serves as a source of iron and is taken up and degraded by a haem oxygenase to liberate the bound iron. We thus tested HemW for haem oxygenase activity in vitro using a method described by Lee et al. [23]. No haem-degrading activity could be detected under a range of conditions, including in the presence of NAD(P)H, SAM (S-adenosylmethionine), Triton X-100 and L. lactis crude extract (results not shown). As an alternative role for HemW, we thus considered a function as a haem chaperone. Indeed, when the HemW–haem complex was incubated with L. lactis membrane vesicles, isolated from cells grown aerobically in the absence of haem, haem was released from the HemW–haem complex (Figure 5). This reaction required NADH and was not observed in the presence of E. coli membrane vesicles used as a control. The release of covalently bound haem from the HemW–haem complex in the presence of L. lactis membranes suggests that haem-loaded HemW can serve in the delivery of haem to a membrane-localized target protein. This target was not cytochrome bd oxidase, as a ΔhemW mutant grown in the presence of haem expressed wild-type levels of cytochrome bd and had no apparent phenotype under standard laboratory growth conditions (results not shown). Future work will have to focus on the identification of the target to which haem from the HemW–haem complex is delivered.

In vitro haem-discharge from HemW

Figure 5
In vitro haem-discharge from HemW

Pre-formed HemW–haem (5 μM) was incubated without (control) or with 1 mg/ml of membrane vesicles from L. lactis (Lla MV) or E. coli (Eco MV). NADH (0.5 mM) was added as indicated and the mixtures were incubated for 1 h at 30°C. Equivalents of 50 pmol of HemW were then resolved by SDS/PAGE, blotted on to nitrocellulose membranes and stained for haem and protein as indicated.

Figure 5
In vitro haem-discharge from HemW

Pre-formed HemW–haem (5 μM) was incubated without (control) or with 1 mg/ml of membrane vesicles from L. lactis (Lla MV) or E. coli (Eco MV). NADH (0.5 mM) was added as indicated and the mixtures were incubated for 1 h at 30°C. Equivalents of 50 pmol of HemW were then resolved by SDS/PAGE, blotted on to nitrocellulose membranes and stained for haem and protein as indicated.

HemW-like proteins form a distinct phylogenetic clade.

HemW-like proteins have previously been called HemN, HemN1, HemZ or YggW and have been assigned to the family of CPDH enzymes. We clearly showed that HemW of L. lactis is not a CPDH, but a haem-binding protein that may act as a haem chaperone. Phylogenetic analysis supports this conclusion: HemW- and HemN-like proteins form two distinct clades (Figure 6). This apparently reflects the different functions of these two protein families. This is further underlined by structural differences. First, HemW-like proteins are significantly shorter than HemN-like proteins and lack N-terminal amino acid residues that have been shown to be functionally relevant for CPDH activity [17]. Secondly, HemW-like proteins feature conserved amino acid residues in the C-terminal portion that are not conserved in HemN-like proteins, namely His134, H184VxxYxLxLE, Y234ExS and especially H248NxxYW, which is always present in HemW-like proteins, but never present in HemN-like proteins. In fact, these residues may serve as sequence flags for the identification of HemW-like proteins. Taken together, in the present study we have identified a novel functional class of haem-binding proteins. Future work will have to show the detailed molecular action of these HemW-like proteins in bacteria.

Phylogram of HemW-like and HemN-like proteins

Figure 6
Phylogram of HemW-like and HemN-like proteins

The phylogram was generated using the ‘one-click’ method with default settings (http://www.Phylogeny.fr) [31]. The scale indicates the branch length for 0.3 nt substitutions per site and the broken line shows the division between the HemW and the HemN clade. The branch labels list the species name, the gene name if available and the GenBank® accession number. Organisms marked with an asterisk feature more than one HemW-like protein. B. japonicum, marked with two asterisks, features two HemN and two HemW-like proteins.

Figure 6
Phylogram of HemW-like and HemN-like proteins

The phylogram was generated using the ‘one-click’ method with default settings (http://www.Phylogeny.fr) [31]. The scale indicates the branch length for 0.3 nt substitutions per site and the broken line shows the division between the HemW and the HemN clade. The branch labels list the species name, the gene name if available and the GenBank® accession number. Organisms marked with an asterisk feature more than one HemW-like protein. B. japonicum, marked with two asterisks, features two HemN and two HemW-like proteins.

DISCUSSION

In the original annotation of the L. lactis genome, the hemW gene was annotated as hemN and was assumed to encode a CPDH. In the present study we have demonstrated that HemW of L. lactis does not function as CPDH, but can bind haem reversibly. In a phylogenetic analysis, it became clear that HemW and HemN proteins form two distinct phylogenetic clades (Figure 6), reflecting the different functions of these two protein families. Members of the HemW clade have previously been called HemN, HemN1, HemZ or YggW, and have wrongly been assigned to the family of CPDH enzymes. The close relationship of HemN and HemW proteins does, however, suggest that these two protein families have a common ancestor. Also, HemW function does not appear to be a universal requirement of living cells, but is present in many different organisms across kingdoms.

On the basis of our experimental and bioinformatics findings, we propose a role of HemW-like proteins in haem trafficking. This is also supported by transcriptional and proteomics profiling of a Shewanella oneidensis FUR (ferric uptake regulator) mutant. The expression of three genes, yggW (hemW-homologue), hemR (Ton-dependent haem receptor) and hut (periplasmic haem-binding protein) were up-regulated 3–4-fold in the fur mutant, compared with the wild-type [24]. The observed FUR-dependent co-regulation of yggW with haem homoeostasis and transport genes supports the hypothesis that YggW (HemW) functions in haem homoeostasis. In E. coli, the corresponding hemW-like gene, yggW, is located in the yggSTUVW operon and has also been shown to bind haem (G. Layer and D. Jahn, unpublished work). Interestingly, another gene of the ygg operon, yggT, encodes an orthologue of Chlamydomonas reinhardtii CBB3, which has been shown to be required for cytochrome b6 maturation [25,26]. Thus the E. coli yggSTUVW operon might play an as yet undiscovered role in haem trafficking and/or cytochrome maturation. L. lactis also features an yggT-like gene, ytdD, which has not yet received further attention.

On the basis of the present study, we propose that HemW binds haem at the membrane and releases it in the presence of NADH, probably to supply it to a membrane-localized target, e.g. cytochromes. None of the haem transport systems that have been described for Gram-positive bacteria, like IsdABC, Shp, HtaAB or HmvTUV, were detected in L. lactis. In E. coli, the ‘dipeptide permease’ DppBCDF has been shown to be a haem transporter [27]. L. lactis features a homologous ‘oligopeptide transporter’, encoded by the optBCDF operon [4]. An involvement of the OptBCDF permease in haem transport remains to be demonstrated. Haem entering the cell could be bound by HemW at the inner face of the cytoplasmic membrane, as suggested by the observation that haem-loaded HemW was only detected in the membrane fraction of HemW-producing E. coli cells. Haem-loaded HemW could then transfers haem in an NADH-dependent reaction to a yet unknown target. This haem transfer only required native L. lactis membranes and haem-loaded HemW in vitro, suggesting that no other cytoplasmic components are required. Of course, it cannot be ruled out that other membrane proteins are involved in the process.

We can only speculate on the role of NADH in the release of haem from HemW. NADH could provide electrons for bond breakage during haem release. Haem transfer from L. lactis HemW to its membrane-associated target was specific, as it was only supported by L. lactis membranes, but not by E. coli membranes. Generally, haem insertions into cytochromes are complex processes and are still poorly understood for most cytochromes. Best understood is haem transfer to apo-cytochrome c. Three different machineries for this process are known, namely systems I, II and III. In many Gram-negative bacteria, system I utilizes the membrane-anchored haem chaperone CcmE that transports haem from the haem b transporter subunit CcmC to the putative haem lyase CcmF [20,28,29]. The release of haem b from the haem chaperone CcmE requires additional proteins including CcmB and most likely the hydrolysis of ATP. In Gram-positive bacteria, system II uses a complex of ResB–CcsB in addition to ResC–CcsA (Ccs is cytochrome c biogenesis protein) that binds haem b in the cytoplasm and deliver it to the extracytoplasmic side for integration into apo-cytochromes [30].

In summary, in the present study we describe a novel haem-binding protein, HemW of L. lactis, with a putative function in haem transfer. HemW belongs to a newly defined family of proteins that occur widely in nature. Clearly, much additional work is required for a detailed understanding of the molecular function of HemW in haem trafficking and investigations in this regard are currently underway in our laboratories.

Abbreviations

     
  • Ccm

    cytochrome c maturation protein

  •  
  • Ccs

    cytochrome c biogenesis protein

  •  
  • CPDH

    coproporphyrinogen III dehydrogenase

  •  
  • Fe–S

    iron–sulfur

  •  
  • FUR

    ferric uptake regulator

  •  
  • IPTG

    isopropyl-β-D-thiogalactoside

  •  
  • LB

    Luria–Bertani

  •  
  • rTEV

    recombinant tobacco edge virus

  •  
  • SAM

    S-adenosylmethionine

AUTHOR CONTRIBUTION

Helge Abicht carried out most of the experiments. Jacobo Martinez performed the experiments in Figure 4. Marc Solioz and Helge Abicht designed the study and wrote the paper. Gunhild Layer and Dieter Jahn analysed results and contributed to the writing of the paper. All authors commented on the paper.

We thank Manfred Heller for performing the MS and Wolfram Saenger for critical reading of the paper. We are grateful to the National BioResource Project (NIG, Japan) for providing strain JW3838.

FUNDING

This work was supported by the Swiss National Foundation [grant number 3100A0–122551] and the International Copper Association. H.K.A. was supported by an EMBO short-term fellowship. G.L. and D.J. were supported by the Deutsche Forschungsgemeinschaft [grant Forschergruppe PROTRAIN, FOR 1220].

References

References
1
Duwat
P.
Sourice
S.
Cesselin
B.
Lamberet
G.
Vido
K.
Gaudu
P.
Le Loir
Y.
Violet
F.
Loubiere
P.
Gruss
A.
Respiration capacity of the fermenting bacterium Lactococcus lactis and its positive effects on growth and survival
J. Bacteriol.
2001
, vol. 
183
 (pg. 
4509
-
4516
)
2
Gaudu
P.
Vido
K.
Cesselin
B.
Kulakauskas
S.
Tremblay
J.
Rezaiki
L.
Lamberret
G.
Sourice
S.
Duwat
P.
Gruss
A.
Respiration capacity and consequences in Lactococcus lactis
Antonie Van Leeuwenhoek
2002
, vol. 
82
 (pg. 
263
-
269
)
3
Sijpesteijn
A. K.
Induction of cytochrome formation and stimulation of oxidative dissimilation by hemin in Streptococcus lactis and Leuconostoc mesenteroides
Antonie Van Leeuwenhoek
1970
, vol. 
36
 (pg. 
335
-
348
)
4
Bolotin
A.
Wincker
P.
Mauger
S.
Jaillon
O.
Malarme
K.
Weissenbach
J.
Ehrlich
S. D.
Sorokin
A.
The complete genome sequence of the lactic acid bacterium Lactococcus lactis ssp. lactis IL1403
Genome Res.
2001
, vol. 
11
 (pg. 
731
-
753
)
5
Xu
K.
Delling
J.
Elliott
T.
The genes required for heme synthesis in Salmonella Typhimurium include those encoding alternative functions for aerobic and anaerobic coproporphyrinogen oxidation
J. Bacteriol.
1992
, vol. 
174
 (pg. 
3953
-
3963
)
6
Xu
K.
Elliott
T.
Cloning, DNA sequence, and complementation analysis of the Salmonella Typhimurium hemN gene encoding a putative oxygen-independent coproporphyrinogen III oxidase
J. Bacteriol.
1994
, vol. 
176
 (pg. 
3196
-
3203
)
7
Layer
G.
Verfurth
K.
Mahlitz
E.
Jahn
D.
Oxygen-independent coproporphyrinogen-III oxidase HemN from Escherichia coli
J. Biol. Chem.
2002
, vol. 
277
 (pg. 
34136
-
34142
)
8
Goto
T.
Aoki
R.
Minamizaki
K.
Fujita
Y.
Functional differentiation of two analogous coproporphyrinogen III oxidases for heme and chlorophyll biosynthesis pathways in the cyanobacterium Synechocystis sp. PCC 6803
Plant Cell Physiol.
2010
, vol. 
51
 (pg. 
650
-
663
)
9
Hippler
B.
Homuth
G.
Hoffmann
T.
Hungerer
C.
Schumann
W.
Jahn
D.
Characterization of Bacillus subtilis hemN
J. Bacteriol.
1997
, vol. 
179
 (pg. 
7181
-
7185
)
10
Homuth
G.
Rompf
A.
Schumann
W.
Jahn
D.
Transcriptional control of Bacillus subtilis hemN and hemZ
J. Bacteriol.
1999
, vol. 
181
 (pg. 
5922
-
5929
)
11
Terzaghi
B. E.
Sandine
W. E.
Improved medium for lactic streptococci and their bacteriophages
Appl. Microbiol.
1975
, vol. 
29
 (pg. 
807
-
813
)
12
Beinert
H.
Micro methods for the quantitative determination of iron and copper in biological material
Methods Enzymol.
1978
, vol. 
54
 (pg. 
435
-
445
)
13
Tsubaki
M.
Nakayama
M.
Okuyama
E.
Ichikawa
Y.
Hori
H.
Existence of two heme B centers in cytochrome b561 from bovine adrenal chromaffin vesicles as revealed by a new purification procedure and EPR spectroscopy
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
23206
-
23210
)
14
Laemmli
U. K.
Favre
M.
Maturation of the head of bacteriophage T4
J. Biol. Chem.
1973
, vol. 
80
 (pg. 
575
-
599
)
15
Feissner
R.
Xiang
Y.
Kranz
R. G.
Chemiluminescent-based methods to detect subpicomole levels of c-type cytochromes
Anal. Biochem.
2003
, vol. 
315
 (pg. 
90
-
94
)
16
Tedesco
A. C.
Oliveira
D. M.
Lacava
Z. G. M.
Azevedo
R. B.
Lima
E. D. C.
Morais
P. C.
Investigation of the binding constant and stoichiometry of biocompatible cobalt ferrite-based magnetic fluids to serum albumin
J. Magn. Magn. Mater.
2004
, vol. 
272–276
 (pg. 
2404
-
2405
)
17
Layer
G.
Moser
J.
Heinz
D. W.
Jahn
D.
Schubert
W. D.
Crystal structure of coproporphyrinogen III oxidase reveals cofactor geometry of Radical SAM enzymes
EMBO J.
2003
, vol. 
22
 (pg. 
6214
-
6224
)
18
Smith
J. M.
Quirk
A. V.
Plank
R. W.
Diffin
F. M.
Ford
G. C.
Harrison
P. M.
The identity of Escherichia coli bacterioferritin and cytochrome b1
Biochem. J.
1988
, vol. 
255
 (pg. 
737
-
740
)
19
Mogi
T.
Miyoshi
H.
Properties of cytochrome bd plastoquinol oxidase from the cyanobacterium Synechocystis sp. PCC 6803
J. Biochem.
2009
, vol. 
145
 (pg. 
395
-
401
)
20
Schulz
H.
Hennecke
H.
Thöny-Meyer
L.
Prototype of a heme chaperone essential for cytochrome c maturation
Science
1998
, vol. 
281
 (pg. 
1197
-
1200
)
21
Daltrop
O.
Stevens
J. M.
Higham
C. W.
Ferguson
S. J.
The CcmE protein of the c-type cytochrome biogenesis system: unusual in vitro heme incorporation into apo-CcmE and transfer from holo-CcmE to apocytochrome
Proc. Natl. Acad. Sci. U.S.A.
2002
, vol. 
99
 (pg. 
9703
-
9708
)
22
Richard-Fogal
C. L.
Frawley
E. R.
Bonner
E. R.
Zhu
H.
San Francisco
B.
Kranz
R. G.
A conserved haem redox and trafficking pathway for cofactor attachment
EMBO J.
2009
, vol. 
28
 (pg. 
2349
-
2359
)
23
Lee
W. C.
Reniere
M. L.
Skaar
E. P.
Murphy
M. E.
Ruffling of metalloporphyrins bound to IsdG and IsdI, two heme-degrading enzymes in Staphylococcus aureus
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
30957
-
30963
)
24
Thompson
D. K.
Beliaev
A. S.
Giometti
C. S.
Tollaksen
S. L.
Khare
T.
Lies
D. P.
Nealson
K. H.
Lim
H.
Yates
J.
III
Brandt
C. C.
, et al. 
Transcriptional and proteomic analysis of a ferric uptake regulator (fur) mutant of Shewanella oneidensis: possible involvement of fur in energy metabolism, transcriptional regulation, and oxidative stress
Appl. Environ. Microbiol.
2002
, vol. 
68
 (pg. 
881
-
892
)
25
Kuras
R.
de Vitry
C.
Choquet
Y.
Girard-Bascou
J.
Culler
D.
Buschlen
S.
Merchant
S.
Wollman
F. A.
Molecular genetic identification of a pathway for heme binding to cytochrome b6
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
32427
-
32435
)
26
Kuras
R.
Saint-Marcoux
D.
Wollman
F. A.
de Vitry
C.
A specific c-type cytochrome maturation system is required for oxygenic photosynthesis
Proc. Natl. Acad. Sci. U.S.A.
2007
, vol. 
104
 (pg. 
9906
-
9910
)
27
Létoffé
S.
Delepelaire
P.
Wandersman
C.
The housekeeping dipeptide permease is the Escherichia coli heme transporter and functions with two optional peptide binding proteins
Proc. Natl. Acad. Sci. U.S.A.
2006
, vol. 
103
 (pg. 
12891
-
12896
)
28
Sanders
C.
Turkarslan
S.
Lee
D. W.
Daldal
F.
Cytochrome c biogenesis: the Ccm system
Trends Microbiol.
2010
, vol. 
18
 (pg. 
266
-
274
)
29
Bonnard
G.
Corvest
V.
Meyer
E. H.
Hamel
P. P.
Redox processes controlling the biogenesis of c-type cytochromes
Antioxid. Redox. Signal.
2010
, vol. 
13
 (pg. 
1385
-
1401
)
30
Ahuja
U.
Kjelgaard
P.
Schulz
B. L.
Thöny-Meyer
L.
Hederstedt
L.
Haem-delivery proteins in cytochrome c maturation system II
Mol. Microbiol.
2009
, vol. 
73
 (pg. 
1058
-
1071
)
31
Dereeper
A.
Guignon
V.
Blanc
G.
Audic
S.
Buffet
S.
Chevenet
F.
Dufayard
J. F.
Guindon
S.
Lefort
V.
Lescot
M.
, et al. 
Phylogeny.fr: robust phylogenetic analysis for the non-specialist
Nucleic Acids Res.
2008
, vol. 
36
 (pg. 
W465
-
W469
)

Supplementary data