A growing family of cytochrome b 5-domain fusion proteins

Abstract

The discovery of cytochrome b5-fused desaturases and hydroxylases (see Fig. 1Fig. 1) has interesting implications for both the evolution of multifunctional, multidomain enzymes and the specificity (i.e. stereochemistry, regioselectivity and substrate specificity) of desaturases. Southern blot analysis of the tobacco genome indicates that cytochrome b5 is present as a small gene family27xTobacco cytochrome b5: cDNA isolation, expression analysis, in vitro protein targeting. Smith, M.A. et al. Plant Mol. Biol. 1994; 25: 527–537Crossref | PubMed | Scopus (26)See all References27. However, plants such as B. officinalis express at least four distinct cytochrome b5-domain proteins: the ‘free’ cytochrome b5 protein, an internal domain of nitrate reductase and two different N-terminal sequences in the Δ6-acyl group- and the Δ8-sphingolipid desaturases. All of these cytochrome b5-domain sequences have significantly diverged, although without modification of the His-Pro-Gly-Gly motif involved in heme-binding9xThe cytochrome b5-fold: an adaptable molecule. Lederer, F. Biochimie. 1994; 76: 674–692Crossref | PubMed | Scopus (79)See all References, 15xA new class of cytochrome b5 fusion proteins. Napier, J.A. et al. Biochem. J. 1997; 328: 717–720PubMedSee all References. It may be that the fusion of the mobile electron carrier (cytochrome b5) to various positions on an acceptor protein provides a kinetic advantage. However, cytochrome b5 and its reductase are usually found in excess in microsomal membranes28xHigh oleic sunflower: studies on composition, desaturation of acyl groups in different lipids, organs. Sperling, P. et al. Z. Naturforsch. 1990; C45: 166–172See all References28, although this may imply that they are not particularly efficient. Another intriguing aspect, is why acyl-desaturases with Δ12- and Δ15-regioselectivities appear to lack a fused cytochrome b5 domain, even though these desaturases are much more prevalent in the plant kingdom. It may be that the presence of the b5-domain is restricted to enzymes that modify the proximal portion of lipid components facing the membrane surface (Δ2-Δ9 in acyl-CoA; O-acyl groups of glycerolipids; N-acyl groups and long-chain bases of ceramides).Fig. 1Examples of the cytochrome b5 superfamily. The position of the diagnostic cytochrome b5 heme-binding domain is indicated, although the exact positions are not to scale. The GenBank accession numbers for the sequences are: Oryza sativa cytochrome b5, X75670; Rattus norvegicus sulphite oxidase, L05084; Saccharomyces cerevisiae cytochrome b2, X03215; Lycopersicon esculentum nitrate reductase, X14060; S. cerevisiae Δ9-fatty acid desaturase (OLE1), J05676; Mortierella alpina Δ5-fatty acid desaturase, AF054824; Borago officinalis Δ6-fatty acid desaturase, U79010; Arabidopsis thaliana Δ8-sphingolipid-desaturase, AJ224161; S. cerevisiae sphingolipid-hydroxylase (FAH1), Z49260; Physcomitrella patens Δ6-fatty acid desaturase AJ222980.View Large Image | Download PowerPoint SlideThese features also call into question whether cytochrome b5 was independently fused to desaturases that had already acquired their different specificities, or whether an ancestral fusion protein for proximal lipid modification duplicated and subsequently evolved into different desaturase/hydroxylase enzymes. Phylogenetic analysis indicates that these fusion events may have happened independently at least twice, with one branch comprising the Δ5-, Δ6- and Δ8-glycerolipid/sphingolipid- desaturases (Fig. 2Fig. 2). With regard to the possible ancestral gene for this branch, it is interesting to note that Δ8-unsaturated long-chain bases are much more wide-spread in present day plants than Δ5- or Δ6-unsaturated fatty acids; this may imply that the sphingolipid-desaturase evolved first. It will be interesting to assess whether there is any change in the fitness of a plant in which the capacity to perform Δ8-sphingolipid-desaturation has been disrupted.Fig.2Phylogenetic tree analysis of cytochrome b5-fusion proteins involved in proximal fatty acid modification. Alignments were generated by the CLUSTAL-X program, and the phylogenetic tree was made with ‘TreeView’. Sequences analysed are: d9Cm=Cyanidioschyzon merolae (AB006677); d9Sc=Saccharomyces cerevisiae Δ9-fatty acid desaturase (OLE1); d5Ma=Mortierella alpina Δ5-fatty acid desaturase; d6Pp=Physcomitrella patens Δ6-fatty acid desaturase; d6Ce=Caenorhabditis elegans Δ6-fatty acid desaturase (AF031477); d5Ce=C. elegans Δ5-fatty acid desaturase (Z81122); d6Bo=Borago officinalis Δ6-fatty acid desaturase; d8Ha=Helianthus annuus Δ8-sphingolipid-desaturase (X87143); d8Bn=Brassica napus Δ8-sphingolipid-desaturase (AJ224160); d8At=Arabidopsis thaliana Δ8-sphingolipid-desaturase (AJ224161); d2Sc=S. cerevisiae sphingolipid-hydroxylase (FAH1). Accession numbers for sequences not described in Fig. 1Fig. 1 are also included.View Large Image | Download PowerPoint Slide