Table 2. Bioinformatic Predictions and in Vivo Validation of Chlamydial Putative Effectors
ProteinBPBAacSIEVEPETSEffective T3Secretion Assay in Shigella
P. amoebophila GlgANot tested (highly conserved)
P. acanthamoeba GlgA+
P. amoebophila GlgB+
P. acanthamoeba GlgB+-
P. amoebophila GlgCN/AN/AN/AN/AN/A
P. acanthamoeba GlgC+++
P. amoebophila GlgPNot tested
P. acanthamoeba GlgP-1+Chimera not expressed
P. acanthamoeba GlgP-2+
P. amoebophila GlgX+++++
P. acanthamoeba GlgXN/AN/AN/AN/AN/A
P. amoebophila MalQ+++
P. acanthamoeba MalQ
  • The different putative effector prediction programs were used respectively with a cutoff of 0.4 (PETS), a SIEVE percentage score above 1, and an effector T3% score above 90% (Arnold et al., 2009; Löwer and Schneider, 2009; Samudrala et al., 2009; Wang et al., 2010). N/A refers to the presence of too many uncertainties on the translation start site to allow prediction or in vivo testing. We were unable to test GlgX from P. acanthamoeba and GlgC from P. amoebophila because of uncertainties in the reported N-terminal sequence deduced from the genome sequence. Because the N-terminal sequence of GlgA from P. acanthamoeba and Protochlamydia was highly conserved, we chose to test only the Parachlamydia sequence. GlgA, glycogen synthase; GlgB, glycogen branching enzyme; GlgC, ADP-Glc pyrophosphorylase; GlgX, debranching enzyme (isoamylase-like); MalQ, α-1,4 glucanotransferase (DPE2 maltase type).