Principal Component Analysis of H8-loop-H9 motifs in bacterial GluRS
Analysis of the H8-L -H9 motif in the bacterial GluRS sequence database was performed using Principal Component Analysis( PCA). All H8-L -H9 sequences from bacterial GluRSs are shown projected on the PC1-PC2 plane (Figure 3A), where PC1 and PC2 correspond to collective sequence-axes associated with maximum mean square fluctuations. The H8-L -H9 sequences clustered broadly into three groups: (A) proteobacterial GluRSs that are incapable of glutamylating tRNAGln, (B) proteobacterial GluRSs that are capable of glutamylating tRNAGln, and, (C) all non-proteobacterial GluRS, irrespective of whether or not they can glutamylate tRNAGln. The PC2 axis separated the proteobacterial GluRSs (groups A & B) from non-proteobacterial GluRSs (group C), indicating that the sequence signature of the H8-L -H9 motif is distinctly different between proteobacterial and non-proteobacterial GluRSs. On the other hand, the PC1 axis separated the proteobacterial GluRSs depending on their tRNAGln-specificity (groups A and B), indicating that the H8-L -H9 motif is distinctly different between tRNAGln-discriminatory GluRSs (D-GluRS/T1-GluRS) and tRNAGln-non-discriminatory GluRSs (ND-GluRS/T2-GluRS).