Principal component analysis of
(𝛔2-𝛔4) - SnoaL2 correlated motions
We also carried out the Principal Component Analysis (PCA), i.e.,
essential dynamics of the covariance matrix resulting from the 200ns MD
trajectories, to investigate the underlying interactions at the NPDKV
and DGGGR motifs and their impact on the overall domain motions.
Principal components (PC) analysis was applied to the backbone atoms of
the RpoE10 and RpoE10 (Mut1, Mut2, and Del1) forms. PC analysis is used
to reveal the functionally relevant and most dominant internal modes of
motion of a MD simulation 35,36. We chose PC1, the
most crucial component, accounting for maximum variability in protein
conformation and computed eigenvectors and associated eigenvalues
(variance). Covariance values provide information about the correlated
motion. All diagonal elements of the covariance matrix were summed and
termed as trace values, which provide information about the measure of
the total variance. As shown in the scatter plot of PC1 vs. PC2 (Figure.
7A-D), the RpoE10 (Del1) and RpoE10 (Mut1) occupy larger subspaces
corresponding to their higher trace values of the covariance matrix
63.03 (nm2) and 58.51 (nm2),
respectively, as compared to RpoE10 (37.16 nm2) and
RpoE10(Mut-2) (43.58 nm2) (Figure. 7(A-D) ). The
higher trace values of RpoE(Del1) and RpoE10(Mut1) relative to the
wild-type RpoE10 and RpoE10(Mut2) suggested an association with an
enhanced flexible behavior upon mutation in NPDKV or the deletion of the
proximal SnoaL2 domain. We observed that the cumulative variance
captured by the first 20 eigenvectors of RpoE10 (wild type) is lower as
compared to the RpoE10(Mut1) and RpoE10(Del1) (Figure 7E). The RpoE10
(Mut1) and RpoE (Del1) showed 72% and 80% variance, respectively, of
the cumulative proportion of the total variance captured by the first
five eigenvectors (Supplemental Table S4). This suggests that the
mutation at the NPDKV motif has impacted the RpoE10-SnoaL2 correlated
motions. The first two principal components PC1 and PC2 account for
~55% and ~64% of the total variance of
all the motions for RpoE10 (Mut1) and RpoE10(Del1), respectively, as
compared to ~ 49% in RpoE10 and 50% in RpoE10(Mut2) (
Supplemental Table S4). The high eigenvalues, i.e., the variance of the
covariance matrix in both RpoE10 (Mut1) and RpoE10 (Del1), indicate the
signals for critical transitions in the conformational changes lead to
enhanced activation. Percentages of variance against eigenvalues of the
covariance matrix resulting from simulations are shown in Figure 7E.
Furthermore, we correlated the elimination or activation of the promoter
with the global protein motion in RpoE10 and RpoE10 (Mut1 and Mut2) and
RpoE10 (Del1) using PCA analysis (Figure 7F, G, H, and I). The mutation
at NPDKV and DGGGR greatly influences the SnoaL_2 domain and overall
dynamics of 2-𝝈4 domain, and magnifies the significant
conformational movements (Figure 7H and I). The PCA indicated that the
essential motion of RpoE10(Mut1) and RpoE10(Del1) was dominated by
fluctuations of the critical L3 loop (residues 46-52), also termed as
“flexibility loop” and Arginine Tetrad (residues 73-76) (marked by a
dotted circle and double arrow and an in Figure 7F, G, H and I) of the
𝛔2 domain. Another noticeable and differential key
fluctuation was observed in the orientation of helix-α7 of
𝛔4 domain (marked by a dotted square in Figure 7 F, G,
H, and I). The helix-α7 constitutes a helix-turn-helix motif that
recognizes the -35 element of the promoter 34 (Figure
4A). The principal differential movements of the backbone atoms at -10
and -35 recognition and residues involved in promoter recognition was
further recorded from the average structures of RpoE10 and its variant
obtained from the PC1 (Figure S2).
A close inspection at -10 promoter binding cleft revealed correlated
motions in RpoE10(Mut1) and RpoE10 (Del1); however, RpoE10 and RpoE10
(Mut2) forms exhibited a significant anti-correlated motion (Figure 7
(bottom panel) and Figure S2A and B). More minor fluctuations in RpoE10,
as well as RpoE10(Mut2) in comparison to RpoE10 (Mut1) and RpoE10(Del1)
were observed in the specificity loop, L3- connecting α2 and α3 (marked
with a circle in Figure 7 (F,G,H, and I)). This differential flexibility
of the L3-loop and orientation of the helix may directly affect the
binding to the promoter region and thus impact the promoter activity.
Notably, the wide-open cleft of -10 recognition site with reduced
flexibility of L3-loop (shown by the blue double arrow in Figure 7) in
both RpoE10 and RpoE10 (Mut2) may constrain the
𝛔2-𝛔4 domain in an “open” and
unproductive conformation leading to the elimination of promoter
activity (Figure 7 ). Another contrasting movement both in the direction
and distance was observed at the WLPEP motif of RpoE10 (Mut1) and
RpoE10(Del1) as compared to RpoE10 and RpoE10(Mut2) (Figure 7, bottom
panel). We noticed an increase in the distance between the initial and
final conformation of Cα for W83 and P87 in RpoE10 (Mut1) and RpoE10
(Del1) as compared to RpoE10 and RpoE10(Mut2) (Figure 7 (bottom panel)
and Supplemental Table S5.) Intriguingly, we find that WLPEP motif moves
inward-up in both RpoE10 and RpoE10(Mut2), whereas in RpoE10(Mut1) and
RpoE10(Del1), the movements are outward-up and outward-down,
respectively. This differential inward and outward trajectories of the
WLPEP motif suggest that it may act as a sensor in transmitting the
conformational signal from the Snoal_2 domain to2-𝛔4domains.