Fig. 4

Domain architectures found in GDPD-like SMaseD/PLD homologs, predominantly in basal and Aquatic clade sequences. a A basal hypothetical protein sequence from Pyrenochaeta lysopersici includes a C-terminal PLAT (Polycystin-1, Lipoxygenase, Alpha-Toxin) repeat domain (cd01756), as well as N-terminal bacterial Ig-like domain repeats (BIg) homologous to those found in calcium-dependent bacterial adhesins (PDB IDs 4P99 and 2YN3), b One basal sequence and one AT-like sequence have an N-terminal domain with similarity to PUD-1/PUD-2 from C. elegans (protein upregulated in daf-2 loss of function); PDB ID 4JDE), c Three basal sequences contain one or two C-terminal PLAT repeats, d Several rotifer sequences in the Aquatic clade contain a C-terminal VMO-I domain (vitelline membrane outer layer protein-I; cd00220), e At least 70% of Aquatic clade sequences contain 1–4 repeats of an unclassified domain of unknown function (labeled DUF-B) with 10 conserved cysteine residues (Additional file 3: Figure S2), f two metagenomic sequences in the Aquatic clade with 74% overall identity conserve a ~ 80-residue Cys-rich domain of unknown function (DUF-A) that is also found in several species of eukaryotic marine phytoplankton, g two basal proteobacterial sequences with 46% overall sequence identity conserve an apparent ~ 150-residue N-terminal domain of unknown function (DUF-C). VMO-I and PLAT repeat domains were identified using CD-search on the NCBI Conserved Domain Database, while PUD and BIg domains were identified using FFAS (see Materials and Methods). Signal peptides were not evident on all sequences; in some but not all cases this may be due to an incomplete N terminal sequence