Limits...
Classification and substrate head-group specificity of membrane fatty acid desaturases

View Article: PubMed Central - PubMed

ABSTRACT

Membrane fatty acid desaturases are a diverse superfamily of enzymes that catalyze the introduction of double bonds into fatty acids. They are essential in a range of metabolic processes, such as the production of omega-3 fatty acids. However, our structure–function understanding of this superfamily is still developing and their range of activities and substrate specificities are broad, and often overlapping, which has made their systematic characterization challenging. A central issue with characterizing these proteins has been the lack of a structural model, which has been overcome with the recent publication of the crystal structures of two mammalian fatty acid desaturases. In this work, we have used sequence similarity networks to investigate the similarity among over 5000 related membrane fatty acid desaturase sequences, leading to a detailed classification of the superfamily, families and subfamilies with regard to their function and substrate head-group specificity. This work will facilitate rapid prediction of the function and specificity of new and existing sequences, as well as forming a basis for future efforts to manipulate the substrate specificity of these proteins for biotechnology applications.

No MeSH data available.


Overview of the sequence similarity relationships in the membrane-bound desaturase family. These representative networks show 2878 nodes representing the 5245 proteins in the membrane-bound desaturase family, in the Pfam database (pfam.xfam.org) (the list of the UniProt IDs of the sequences included in this network are provided in Supplementary Table S1). The clusters with functionally characterized member sequences are labelled, including the functionally dissimilar alkane monooxygenases cluster and beta-carotene ketolases clusters. There are three major fatty acid desaturase families (first desaturases, FDs; methyl-end desaturases, MEs; front-end desaturases, FEs). Δ4 sphingolipid desaturases (SDs) are also shown. Larger squares represent nodes with at least one functionally characterized member. Edges or lines connecting the nodes are shown if the pairwise similarity score between the sequences of representative nodes is lower than the threshold of Log BLAST E-value of − 13. The 366,864 edges in the overview network have a median sequence identity of 30% over 275 residues. A and B are identical networks with different colour coding as explained in the figure keys. A. The nodes in this overview network are coloured by the kingdom of the organism of origin. B. The nodes in this overview network are coloured by the substrate head-group specificity of the characterized members.Overview of the sequence similarity relationships in the membrane-bound desaturase family. These representative networks show 2878 nodes representing the 5245 proteins in the membrane-bound desaturase family, in the Pfam database (pfam.xfam.org) (the list of the UniProt IDs of the sequences included in this network are provided in Supplementary Table S1). The clusters with functionally characterized member sequences are labelled, including the functionally dissimilar alkane monooxygenases cluster and beta-carotene ketolases clusters. There are three major fatty acid desaturase families (first desaturases, FDs; methyl-end desaturases, MEs; front-end desaturases, FEs). Δ4 sphingolipid desaturases (SDs) are also shown. Larger squares represent nodes with at least one functionally characterized member. Edges or lines connecting the nodes are shown if the pairwise similarity score between the sequences of representative nodes is lower than the threshold of Log BLAST E-value of − 13. The 366,864 edges in the overview network have a median sequence identity of 30% over 275 residues. A and B are identical networks with different colour coding as explained in the figure keys. A. The nodes in this overview network are coloured by the kingdom of the organism of origin. B. The nodes in this overview network are coloured by the substrate head-group specificity of the characterized members.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5037126&req=5

f0010: Overview of the sequence similarity relationships in the membrane-bound desaturase family. These representative networks show 2878 nodes representing the 5245 proteins in the membrane-bound desaturase family, in the Pfam database (pfam.xfam.org) (the list of the UniProt IDs of the sequences included in this network are provided in Supplementary Table S1). The clusters with functionally characterized member sequences are labelled, including the functionally dissimilar alkane monooxygenases cluster and beta-carotene ketolases clusters. There are three major fatty acid desaturase families (first desaturases, FDs; methyl-end desaturases, MEs; front-end desaturases, FEs). Δ4 sphingolipid desaturases (SDs) are also shown. Larger squares represent nodes with at least one functionally characterized member. Edges or lines connecting the nodes are shown if the pairwise similarity score between the sequences of representative nodes is lower than the threshold of Log BLAST E-value of − 13. The 366,864 edges in the overview network have a median sequence identity of 30% over 275 residues. A and B are identical networks with different colour coding as explained in the figure keys. A. The nodes in this overview network are coloured by the kingdom of the organism of origin. B. The nodes in this overview network are coloured by the substrate head-group specificity of the characterized members.Overview of the sequence similarity relationships in the membrane-bound desaturase family. These representative networks show 2878 nodes representing the 5245 proteins in the membrane-bound desaturase family, in the Pfam database (pfam.xfam.org) (the list of the UniProt IDs of the sequences included in this network are provided in Supplementary Table S1). The clusters with functionally characterized member sequences are labelled, including the functionally dissimilar alkane monooxygenases cluster and beta-carotene ketolases clusters. There are three major fatty acid desaturase families (first desaturases, FDs; methyl-end desaturases, MEs; front-end desaturases, FEs). Δ4 sphingolipid desaturases (SDs) are also shown. Larger squares represent nodes with at least one functionally characterized member. Edges or lines connecting the nodes are shown if the pairwise similarity score between the sequences of representative nodes is lower than the threshold of Log BLAST E-value of − 13. The 366,864 edges in the overview network have a median sequence identity of 30% over 275 residues. A and B are identical networks with different colour coding as explained in the figure keys. A. The nodes in this overview network are coloured by the kingdom of the organism of origin. B. The nodes in this overview network are coloured by the substrate head-group specificity of the characterized members.

Mentions: In order to characterize the sequence and functional diversity in the membrane FADs, particularly the substrate head-group specificity, 5245 sequences were collected using the PFAM fatty acid desaturase family PF00487 as seed sequence clusters within the length range of 350 aa to 550 aa. This length range was chosen to limit the search to the single domain desaturases and the desaturases with a fused cytochrome b5 domain. The free cytochrome b5 proteins and the unrelated long cytochrome b5 domain containing fusion proteins are eliminated with this filter. The collected sequences were analysed by generating a SSN using EFI-EST [35], where an initial network containing 2878 representative nodes (clusters of protein sequences with > 60% amino acid identity) was produced. The edges represent all-vs-all BLAST E-values between the clusters. The cytochrome b5 domain was only included in the E-value calculations when all of the protein members in the connecting nodes also contained cytochrome b5 domains. Otherwise, only the fatty acid desaturase sequences were used to calculate the E-values. A literature review was performed to collect the known lipid head-group preferences of the characterized desaturases and their organism distributions. This data was mapped onto the networks (Fig. 2).


Classification and substrate head-group specificity of membrane fatty acid desaturases
Overview of the sequence similarity relationships in the membrane-bound desaturase family. These representative networks show 2878 nodes representing the 5245 proteins in the membrane-bound desaturase family, in the Pfam database (pfam.xfam.org) (the list of the UniProt IDs of the sequences included in this network are provided in Supplementary Table S1). The clusters with functionally characterized member sequences are labelled, including the functionally dissimilar alkane monooxygenases cluster and beta-carotene ketolases clusters. There are three major fatty acid desaturase families (first desaturases, FDs; methyl-end desaturases, MEs; front-end desaturases, FEs). Δ4 sphingolipid desaturases (SDs) are also shown. Larger squares represent nodes with at least one functionally characterized member. Edges or lines connecting the nodes are shown if the pairwise similarity score between the sequences of representative nodes is lower than the threshold of Log BLAST E-value of − 13. The 366,864 edges in the overview network have a median sequence identity of 30% over 275 residues. A and B are identical networks with different colour coding as explained in the figure keys. A. The nodes in this overview network are coloured by the kingdom of the organism of origin. B. The nodes in this overview network are coloured by the substrate head-group specificity of the characterized members.Overview of the sequence similarity relationships in the membrane-bound desaturase family. These representative networks show 2878 nodes representing the 5245 proteins in the membrane-bound desaturase family, in the Pfam database (pfam.xfam.org) (the list of the UniProt IDs of the sequences included in this network are provided in Supplementary Table S1). The clusters with functionally characterized member sequences are labelled, including the functionally dissimilar alkane monooxygenases cluster and beta-carotene ketolases clusters. There are three major fatty acid desaturase families (first desaturases, FDs; methyl-end desaturases, MEs; front-end desaturases, FEs). Δ4 sphingolipid desaturases (SDs) are also shown. Larger squares represent nodes with at least one functionally characterized member. Edges or lines connecting the nodes are shown if the pairwise similarity score between the sequences of representative nodes is lower than the threshold of Log BLAST E-value of − 13. The 366,864 edges in the overview network have a median sequence identity of 30% over 275 residues. A and B are identical networks with different colour coding as explained in the figure keys. A. The nodes in this overview network are coloured by the kingdom of the organism of origin. B. The nodes in this overview network are coloured by the substrate head-group specificity of the characterized members.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5037126&req=5

f0010: Overview of the sequence similarity relationships in the membrane-bound desaturase family. These representative networks show 2878 nodes representing the 5245 proteins in the membrane-bound desaturase family, in the Pfam database (pfam.xfam.org) (the list of the UniProt IDs of the sequences included in this network are provided in Supplementary Table S1). The clusters with functionally characterized member sequences are labelled, including the functionally dissimilar alkane monooxygenases cluster and beta-carotene ketolases clusters. There are three major fatty acid desaturase families (first desaturases, FDs; methyl-end desaturases, MEs; front-end desaturases, FEs). Δ4 sphingolipid desaturases (SDs) are also shown. Larger squares represent nodes with at least one functionally characterized member. Edges or lines connecting the nodes are shown if the pairwise similarity score between the sequences of representative nodes is lower than the threshold of Log BLAST E-value of − 13. The 366,864 edges in the overview network have a median sequence identity of 30% over 275 residues. A and B are identical networks with different colour coding as explained in the figure keys. A. The nodes in this overview network are coloured by the kingdom of the organism of origin. B. The nodes in this overview network are coloured by the substrate head-group specificity of the characterized members.Overview of the sequence similarity relationships in the membrane-bound desaturase family. These representative networks show 2878 nodes representing the 5245 proteins in the membrane-bound desaturase family, in the Pfam database (pfam.xfam.org) (the list of the UniProt IDs of the sequences included in this network are provided in Supplementary Table S1). The clusters with functionally characterized member sequences are labelled, including the functionally dissimilar alkane monooxygenases cluster and beta-carotene ketolases clusters. There are three major fatty acid desaturase families (first desaturases, FDs; methyl-end desaturases, MEs; front-end desaturases, FEs). Δ4 sphingolipid desaturases (SDs) are also shown. Larger squares represent nodes with at least one functionally characterized member. Edges or lines connecting the nodes are shown if the pairwise similarity score between the sequences of representative nodes is lower than the threshold of Log BLAST E-value of − 13. The 366,864 edges in the overview network have a median sequence identity of 30% over 275 residues. A and B are identical networks with different colour coding as explained in the figure keys. A. The nodes in this overview network are coloured by the kingdom of the organism of origin. B. The nodes in this overview network are coloured by the substrate head-group specificity of the characterized members.
Mentions: In order to characterize the sequence and functional diversity in the membrane FADs, particularly the substrate head-group specificity, 5245 sequences were collected using the PFAM fatty acid desaturase family PF00487 as seed sequence clusters within the length range of 350 aa to 550 aa. This length range was chosen to limit the search to the single domain desaturases and the desaturases with a fused cytochrome b5 domain. The free cytochrome b5 proteins and the unrelated long cytochrome b5 domain containing fusion proteins are eliminated with this filter. The collected sequences were analysed by generating a SSN using EFI-EST [35], where an initial network containing 2878 representative nodes (clusters of protein sequences with > 60% amino acid identity) was produced. The edges represent all-vs-all BLAST E-values between the clusters. The cytochrome b5 domain was only included in the E-value calculations when all of the protein members in the connecting nodes also contained cytochrome b5 domains. Otherwise, only the fatty acid desaturase sequences were used to calculate the E-values. A literature review was performed to collect the known lipid head-group preferences of the characterized desaturases and their organism distributions. This data was mapped onto the networks (Fig. 2).

View Article: PubMed Central - PubMed

ABSTRACT

Membrane fatty acid desaturases are a diverse superfamily of enzymes that catalyze the introduction of double bonds into fatty acids. They are essential in a range of metabolic processes, such as the production of omega-3 fatty acids. However, our structure–function understanding of this superfamily is still developing and their range of activities and substrate specificities are broad, and often overlapping, which has made their systematic characterization challenging. A central issue with characterizing these proteins has been the lack of a structural model, which has been overcome with the recent publication of the crystal structures of two mammalian fatty acid desaturases. In this work, we have used sequence similarity networks to investigate the similarity among over 5000 related membrane fatty acid desaturase sequences, leading to a detailed classification of the superfamily, families and subfamilies with regard to their function and substrate head-group specificity. This work will facilitate rapid prediction of the function and specificity of new and existing sequences, as well as forming a basis for future efforts to manipulate the substrate specificity of these proteins for biotechnology applications.

No MeSH data available.