Lflvkkpdgtipgyephehggataksgesge P Tggaaahe H E H

Fig. 10. Type I, II, and III copper-site ligands of CAB45584. Amino acid sequence (343 residues) of the putative multicopper oxidase (CAB45584) of S. coelicolor. The ligands of the type I, type II, and type III coppers are indicated by T1, T2, and T3, respectively. The signal peptidase cleavage site as predicted by SignalP (62) is indicated by an asterisk.

The main difference between the oxidases is their substrate, cytochrome-c or quinols. Subunit I contains the redox active copper CuB and binds the two hemes, a and a3. In subunit II, the CuA site is located. The third subunit has the least sequence similarity and its role in the oxidase activity is still a matter of debate. Genes for a terminal oxidase of this family were expected to be present on the Streptomyces coelicolor genome because it has an aerobic metabolism. CAB39882 and CAB39883 show the highest similarity with subunit I and with subunit II of cytochrome-c oxidases, respectively. The two genes are located in a region of the genome encoding several other proteins involved in respiration, including the subunit III of COX.

The other copy of subunit I (CAB94657) is 87% similar to CAB39882 and is similar to many other subunits I of COX and QOX in the database. However, there are no genes for the oxidase subunit II and III located in the upstream or downstream sequences in this region of the genome. It remains to be investigated what the function of this protein is.

4.2. Multicopper Oxidase

The best known representatives of the multicopper oxidases are laccase and ascorbate oxidase (reviewed in ref. 66). These enzymes contain three copper centers, type I, type II, and type III, each characterized by specific absorption bands. Two signatures (Prosite, SIB) in the primary amino acid sequence are indicative for multicopper oxidases. One pattern is specific for copper-binding domains, whereas the other pattern also recognizes domains that have lost their copper-binding ability. In addition, four amino acid sequences containing the copper ligands can be distinguished. Two regions are located in the N-terminal half of the protein and the other two are found in the C-terminal half. Together, these sequences make up the three copper centers.

CAB45586 contains all of the above and, strictly on amino acid similarity, should be considered as a member of the multicopper oxidase family. However, this S. coelicolor protein is much smaller than the known multicopper oxidases, 343 amino acids compared to >600 residues. The four amino acid motifs involved in copper binding are situated closer together in the primary amino acid sequence than is the case in other multicopper oxidases (Fig. 10). The amino acid residues 70-300 show similarity (40-45%) with the 2 domains containing the 4 sequences involved in copper binding of the copper resistance protein A (copA) of Pseudomonas syringae (67), the copper resistance protein

Fig. 11. Multiple-sequence alignment of the S. coelicolor tyrosinase-like protein and all known strepto-mycetes tyrosinases. Alignment of the sequences was carried out with the program MultAlin (75) and shading with the program BoxShade (www.ch.embnet.org/software/). The conserved histidine ligands of CuA and CuB are indicated by an asterisk.

Fig. 11. Multiple-sequence alignment of the S. coelicolor tyrosinase-like protein and all known strepto-mycetes tyrosinases. Alignment of the sequences was carried out with the program MultAlin (75) and shading with the program BoxShade (www.ch.embnet.org/software/). The conserved histidine ligands of CuA and CuB are indicated by an asterisk.

pcoA of Escherichia coli (68), ceruloplasmin, coagulation factor V, laccase, and ascorbate oxidase. At the extreme N-terminal end and at the C-terminal end, no significant similarity shows up in the database.

The first 30 N-terminal amino acids of CAB45586 are predicted to be a signal sequence (62). Therefore, this multicopper oxidase is most likely exported and fulfills its function outside the cell.

4.3. CAB45584

CAB45584 is a protein of 729 amino acids with a putative N-terminal signal sequence. The first 130 amino acid residues of the mature protein have similarity with a putative blue copper protein of Arabidopsis and with auracyanin B of Chloroflexus aurantiacus (69). The blue copper proteins are characterized by a type I copper site with the signature CX1-2P/GX0-iHX2-4M/Q where the C, H, and M are copper ligands. The fourth copper ligand is located many residues upstream of the cysteine. In parts of the 500 N-terminal amino acids, weak similarity with glycosyl hydrolases such as endoglucanase D of Cellumonas fimi is found. However, this class of enzymes has not been reported to contain a copper cofactor. Therefore, a prediction of the enzymatic function of CAB45584 will need more research than just in silico studies to define its function and copper content.

4.4. Tyrosinase-Like Protein

Tyrosinases are enzymes that catalyze the hydroxylation of monophenols to diphenols and the oxidation of diphenols to quinones (reviewed in ref. 70). The latter are precursors of melanines and can spontaneously polymerize to these well-known brown pigments. For the reactions with monophenols, tyrosinases have to bind oxygen. This is accomplished by a dinuclear Cu site (CuA and CuB) present in the enzyme. Three histidine residues bind each copper and together they form the so-called type III site (71,72). Not only do monooxygenases such as tyrosinase have this type of copper-binding site. In hemocyanins, the type III copper site is used to transport and store oxygen (73). The third member of the type III copper-site family is catechol oxidase (74), which can carry out the oxidation of diphenols but not the hydroxylation of monophenols.

Tyrosinases are widespread in nature and occur in all kingdoms. The protein identified in the S. coelicolor genome sequence does contain the conserved histidines involved in copper binding and has similarity with tyrosinases from other Streptomyces strains. However, the similarity is lower than found between the known tyrosinases (Fig. 11). In strains producing tyrosinase, the gene encoding tyrosinase, melC2, is always preceded by a small ORF, melCI, that encodes a histidine-rich protein that is proposed to be the copper chaperone of apotyrosinase and also involved in secretion of tyrosi-nase (76,77). Streptomyces tyrosinase is an extracellular enzyme but does not have an N-terminal signal sequence. This feature is also present in S. coelicolor. Immediately upstream of the gene encoding CAB92266, in fact the start codon of the melC2 ortholog, is located within this ORF; a small ORF is located with many histidine residues and similarity to melCI (Fig. 12). All tyrosinase-expressing Streptomyces strains do produce melanin, but S. coelicolor does not. Taking this into consideration, it suggests that CAB92266 is an extracellular copper enzyme but most likely does not have monophenols as substrate and is not a monophenol monooxygenase. If the protein turns out to be a tyrosinase, its expression must be strongly downregulated under "normal" growth conditions or the enzymatic product cannot polymerize to form melanines.

4.5. CAB50963

This protein consists of 310 amino acids. The C-terminal domain is rich in cysteines (23 out of 68 amino acids) and has a similarity to metallothioneins. The amino-terminal domain does not show significant similarity to proteins in the databases. The protein has been described first for S. albogriseolus as a putative repressor of the expression of an extracellular protease (79). Metallothioneins are small proteins (68 amino acids) that bind heavy metals such as Cu, Zn, Cd, and

Fig. 12. Alignment of the S. coelicolor MelC1 with all known Streptomyces MelC1s. The histidine residues proposed to be involved in copper binding and apotyrosinase maturation are indicated with an & (78). The signal peptidase cleavage site as predicted by SignalP (62) is behind amino acid residue 62, resulting in a abnormally long signal sequence. An alternative translational start site is present (Met at position 35), but this AUG codon is not preceded by an obvious ribosome-binding site, whereas the first methionine codon is.

Fig. 12. Alignment of the S. coelicolor MelC1 with all known Streptomyces MelC1s. The histidine residues proposed to be involved in copper binding and apotyrosinase maturation are indicated with an & (78). The signal peptidase cleavage site as predicted by SignalP (62) is behind amino acid residue 62, resulting in a abnormally long signal sequence. An alternative translational start site is present (Met at position 35), but this AUG codon is not preceded by an obvious ribosome-binding site, whereas the first methionine codon is.

Hg (80). They function as metal sinks in the cell. The appearance of a metallothioneinlike domain in a much larger protein that may be involved in transcription regulation could turn out to be an interesting link with metal homeostasis.

4.6. Metal Transport

The physiological requirement for copper puts the cell in a difficult position. These redox active metals should not be present in the cells in a free state because of their capacity to engage in all kinds of redox reactions that can damage many macromolecules. Therefore, organisms have developed systems to import, sequester, distribute, and export metal ions depending on the intracellular need and the extracellular availability. For the acquisition of metals, the first barrier is the outer membrane. One of the families of proteins taking care of metal transport across membranes are the P-type ATPases containing a heavy-metal-associated domain (HMA) in the N-terminal sequence. The well-studied CopA of Enterococcus hirae belongs to this family (81). The HMA contains a short sequence, GMXCXXC, involved in metal binding. This motif has been used to screen the S. coelicolor genome. Three genes were identified that encode for integral membrane proteins with all characteristics of the heavy-metal pumps (Fig. 13A). They have eight transmembrane domains, a phosphatase domain

Fig. 13. (A) Gene organization of the copper P-type ATPases and copper chaperones. (B) Model for the copper trafficking in S. coelicolor. Two of the P-type ATPases are proposed to be involved in copper uptake, whereas the third ATPase, which is not linked to a chaperone, is concerned with copper efflux in analogy with CopA of E. coli (82). The two copper-uptake ATPases do not necessarily have to be present at the same time in the same membranes. A scenario in which one of them is expressed during vegetative growth and the other during aerial hyphae development could be feasible. The cupric reductase is assumed to be a membrane-bound enzyme because all cupric reductase activity in Streptomyces was found to be mycelium associated (Vijgenboom, unpublished).

Fig. 13. (A) Gene organization of the copper P-type ATPases and copper chaperones. (B) Model for the copper trafficking in S. coelicolor. Two of the P-type ATPases are proposed to be involved in copper uptake, whereas the third ATPase, which is not linked to a chaperone, is concerned with copper efflux in analogy with CopA of E. coli (82). The two copper-uptake ATPases do not necessarily have to be present at the same time in the same membranes. A scenario in which one of them is expressed during vegetative growth and the other during aerial hyphae development could be feasible. The cupric reductase is assumed to be a membrane-bound enzyme because all cupric reductase activity in Streptomyces was found to be mycelium associated (Vijgenboom, unpublished).

(TGE), a conserved phosphorylation site (DKTGT), an ATP-binding domain (GDGVN), and the CPC motif in the sixth transmembrane domain that is conserved in all P-type ATPases involved in transport of the soft metal ions [Cu(I), Cd(II), Ag(I), or Zn(II)]. The N-terminal cytoplasmic domain contains the GMTCAA/SC motif that is indicative for heavy-metal binding. This motif is capable of binding several heavy-metal ions, but the N-terminal domains of ATPases that transport a particular metal ion show a slightly greater similarity (83). This criterion classifies the S. coelicolor P-type

ATPases as copper transporters. The conclusion is supported by the fact that upstream of two of the three ATPases, a small protein is encoded with similarity to CopZ (84,85). This protein is a member of the family of copper chaperones that obtain the copper ion from the ATPase and take care of its distribution in the cell. The copper chaperones contain an N-terminal metal-binding motif very similar to that of the ATPases, GMSCGHC in the case of S. coelicolor. The presence of three putative copper-transporting ATPases in S. coelicolor raises several questions. Are all three ATPases involved in copper uptake? Where are the ATPases located, in the vegetative mycelial membrane or maybe in the septum separating vegetative and aerial hyphae? What are the targets of the copper chaperones? The answers to these questions of course have to await further studies, but an educated guess for two of them can be made right now. The two ATPases that are coexpressed with a copper chaperone are most likely involved in copper uptake. The third ATPase could be concerned with copper export, as has been shown for CopA of E. coli (82). One of the targets for the copper chaperones could be a transcriptional regulator as shown for the CopZ-CopY couple (86) or cytochrome-c oxidase, as shown in Saccharomyces cerevisae (87). A model for copper trafficking is depicted in Fig. 13B. The proposed cupric reductase has not been identified in the S. coelicolor genome sequence, but a cupric-reductase activity is present in intact vegetative mycelium of S. lividans and S. coelicolor (Vijgenboom, unpublished).

0 0

Post a comment